feat(catalogue): one page, one axis — setup summary over per-family engines and weights

The Model Catalogue put the same decision on two axes: an Engines pane with
TTS/ASR/LLM tabs and a Models pane with TTS/ASR/Dictation/Diarisation
sections, dictation shown in both, plus storage stats, the HF token and the
voice-preview toggle parked on the model list. Settings → Voice still carried
Engines and Models entries that only pointed back here.

Now the page reads top-down: a SetupSummary (speech, transcription,
dictation, language model — engine, device, one status word, Change), the
engine list for one family, and that family's downloadable weights under it
(TTS under TTS; offline ASR, streaming dictation and diarisation under ASR;
nothing for LLM, whose engines bring their own). One storage line points at
Settings → Storage.

- ModelStoreTab takes a `family` and scopes sections and the recommendation
  preset to it (scopeReco); stats strip, HF-token toolbar and previews
  panel removed from it.
- Settings: Engines/Models categories and CataloguePointer removed; models
  directory → Storage, HF mirror → Network (both restart-flagged), voice
  previews → Storage. "Manage models" in disk usage opens the catalogue.
- Store: openCatalogue takes a family (pane key tolerated, ignored);
  pendingCatalogueTab gone.
- Engine matrix title is now the locale's plain "Engines".
- i18n: catalogue.* summary keys in all 21 locales; pane/pointer keys dropped.
- Docs: "Model Catalogue → Engines" is "Model Catalogue"; "→ Models" is
  "→ Downloaded weights".

docs/DESKTOP_RELEASE.md

@@ -69,7 +69,7 @@ We exclude every `nvidia.*` wheel from the Apple Silicon build (saves ~2 GB) and
 
 For us:
 - **Base DMG ships MPS + MLX path only.** Excludes `nvidia.*`, `triton`, `flash-attn`, anything CUDA-specific in the spec.
-- **Optional pack: VoxCPM2** (requires CUDA). Not installed by default. Model Catalogue → Engines → "Install VoxCPM2" triggers download from our `voxcpm2-cu128-v1.tar.gz` release asset.
+- **Optional pack: VoxCPM2** (requires CUDA). Not installed by default. Model Catalogue → "Install VoxCPM2" triggers download from our `voxcpm2-cu128-v1.tar.gz` release asset.
 - **Optional pack: pyannote** (HF-token gated). Default off. Settings → Speaker diarisation → "Enable" prompts for HF token, downloads + installs.
 - **Optional pack: MOSS-TTS-Nano.** Same pattern.
 

docs/downloading-models.md

@@ -60,7 +60,7 @@ State is reported at **Settings → About** / `GET /system/info`:
 - `fast_download.xet_installed` — `hf_xet` present (true)
 - `fast_download.xet_active` — whether Xet actually drives downloads (false by
   default, because of `HF_HUB_DISABLE_XET`)
-- the **⚡ fast download** badge in **Model Catalogue → Models** appears only when Xet
+- the **⚡ fast download** badge in **Model Catalogue → Downloaded weights** appears only when Xet
   is *active*.
 
 The backend logs one line at startup, e.g.
@@ -148,7 +148,7 @@ failed download at once. Caveats:
 
 ## Cancelling a download
 
-**Model Catalogue → Models** lets you cancel an in-flight install. Cancellation stops
+**Model Catalogue → Downloaded weights** lets you cancel an in-flight install. Cancellation stops
 further retries and clears the failure cooldown so you can restart
 immediately. A file that's already streaming finishes first — cancellation
 takes effect at the next retry boundary.
@@ -162,6 +162,6 @@ takes effect at the next retry boundary.
   High-performance mode only helps if RAM and bandwidth are plentiful.
 - **"download finished but no model weights were found"** — the download was
   interrupted and left a partial snapshot. Delete the model in
-  **Model Catalogue → Models** and install it again.
+  **Model Catalogue → Downloaded weights** and install it again.
 - **Out of disk** — model sizes are shown in the catalog; free space or change
   the cache location with `HF_HOME` / `HF_HUB_CACHE`.

docs/engines/README.md

@@ -1,7 +1,7 @@
 # Engine guides
 
 One page per engine: what it's for, what it needs, how to enable it, and its
-quirks. Select engines in **Model Catalogue → Engines** (or quick-switch with
+quirks. Select engines in **Model Catalogue** (or quick-switch with
 <kbd>Ctrl</kbd>/<kbd>Cmd</kbd>+<kbd>E</kbd>), or pin one with
 `OMNIVOICE_TTS_BACKEND` / `OMNIVOICE_ASR_BACKEND`.
 

docs/engines/audio-cpp.md

@@ -80,11 +80,11 @@ instead of that dedicated-VRAM floor.
    ```
 
    Alternatively set `OMNIVOICE_AUDIOCPP_DIR` to the directory containing it.
-3. Restart VoiceStudio, open **Model Catalogue → Models**, find
+3. Restart VoiceStudio, open **Model Catalogue → Downloaded weights**, find
    **Breeze-TTS-2 Q8_0 for audio.cpp**, review its research/non-commercial
    license note, and click **Install**. Generation never starts this ~4.73 GiB
    download automatically.
-4. Pick `audiocpp` in **Model Catalogue → Engines**. The server starts
+4. Pick `audiocpp` in **Model Catalogue**. The server starts
    lazily on first generate (`server.json` + `server.log` live under the app
    data `audiocpp/` directory).
 
@@ -131,7 +131,7 @@ The managed loopback port may be taken. Check `server.log` next to
 
 ### `Breeze-TTS-2 ... not installed` or `package ... not completely installed`
 
-Install the model from **Model Catalogue → Models**. If an interrupted install
+Install the model from **Model Catalogue → Downloaded weights**. If an interrupted install
 left it incomplete, use **Reinstall** there. If the error persists after a
 complete reinstall, file an issue with the package listing.
 

docs/engines/confucius4-tts.md

@@ -56,7 +56,7 @@ Then point VoiceStudio at the clone and restart:
 - **macOS/Linux:** `export OMNIVOICE_CONFUCIUS4_TTS_DIR=/path/to/Confucius4-TTS`
 - **Windows (PowerShell):** `[Environment]::SetEnvironmentVariable("OMNIVOICE_CONFUCIUS4_TTS_DIR","C:\path\to\Confucius4-TTS","User")`
 
-Select **Confucius4-TTS** in Model Catalogue → Engines. The first synthesize triggers
+Select **Confucius4-TTS** in Model Catalogue. The first synthesize triggers
 the weight downloads above, then generates.
 
 ### Optional overrides

docs/engines/dots-tts.md

@@ -11,14 +11,14 @@ pins `transformers>=5.3` — the same isolation primitive used by
 [IndexTTS-2](indextts.md) and [MOSS-TTS-v1.5](moss-tts-v15.md).
 
 > **Opt-in, and never a default.** dots.tts is selected explicitly in
-> **Model Catalogue → Engines** (or `OMNIVOICE_TTS_BACKEND=dots-tts`). It is not
+> **Model Catalogue** (or `OMNIVOICE_TTS_BACKEND=dots-tts`). It is not
 > part of the default install.
 
 ## Platform support
 
 - **Linux / macOS only.** dots.tts's upstream package declares Linux and
   macOS classifiers and has **no Windows install path**. On Windows the
-  engine reports itself unavailable in **Model Catalogue → Engines** with a clear
+  engine reports itself unavailable in **Model Catalogue** with a clear
   reason — run VoiceStudio under WSL2 or use a Linux/macOS host.
 - **No MPS.** Upstream device selection is CUDA-or-CPU with no Metal branch,
   so on Apple Silicon the official package runs on **CPU** (slow but
@@ -60,7 +60,7 @@ dots.tts is **not** bundled (large checkpoint + conflicting `transformers`).
    source ~/.zshrc
    ```
 
-5. Restart VoiceStudio. dots.tts appears in **Model Catalogue → Engines** with
+5. Restart VoiceStudio. dots.tts appears in **Model Catalogue** with
    `available: true` and `isolation_mode: subprocess`.
 
 ## Venv resolution order

docs/engines/faster-whisper-isolated.md

@@ -13,7 +13,7 @@ only the process boundary is new.
 
 ## Selecting it
 
-- **Model Catalogue → Engines**, ASR tab → **Use** on the crash-isolated row, or
+- **Model Catalogue**, ASR tab → **Use** on the crash-isolated row, or
 - pin it with `OMNIVOICE_ASR_BACKEND=faster-whisper-isolated`.
 
 It is never picked by auto-detect — it's an explicit opt-in escape hatch.
@@ -56,6 +56,6 @@ Weights download on first load — see
   "sidecar crashed" error and the backend stays up — retry to respawn.
 - **cuDNN 8 is still required on CUDA** — same CTranslate2 requirement as the
   in-process engine. It's checked up front so a missing cuDNN 8 shows as
-  "unavailable" in Model Catalogue → Engines instead of a sidecar that
+  "unavailable" in Model Catalogue instead of a sidecar that
   silently fails every transcribe
   ([#1371](https://github.com/debpalash/VoiceStudio/issues/1371)).

docs/engines/faster-whisper.md

@@ -7,7 +7,7 @@ fallback on non-Apple machines.
 
 ## Selecting it
 
-- **Model Catalogue → Engines**, ASR tab → **Use** on the Faster-Whisper row, or
+- **Model Catalogue**, ASR tab → **Use** on the Faster-Whisper row, or
 - pin it with `OMNIVOICE_ASR_BACKEND=faster-whisper`.
 
 Auto-detect only picks it when [whisperx](whisperx.md) is unavailable.

docs/engines/funasr.md

@@ -10,7 +10,7 @@ WhisperX remains the cross-platform default.
 ## Selecting it
 
 - Install it into the app venv: `uv pip install funasr`.
-- Then **Model Catalogue → Engines**, ASR tab → **Use** on the FunASR row, or
+- Then **Model Catalogue**, ASR tab → **Use** on the FunASR row, or
   `OMNIVOICE_ASR_BACKEND=funasr`.
 
 Auto-detect never picks it; it's an explicit opt-in.

docs/engines/gpt-sovits.md

@@ -24,7 +24,7 @@ HTTP.
    python api_v2.py -a 127.0.0.1 -p 9880 -c GPT_SoVITS/configs/tts_infer.yaml
    ```
 
-2. Select the engine via **Model Catalogue → Engines** or
+2. Select the engine via **Model Catalogue** or
    `OMNIVOICE_TTS_BACKEND=gpt-sovits`.
 
 VoiceStudio marks the engine available only when the server responds

docs/engines/indextts.md

@@ -14,7 +14,7 @@ runtime. Existing user-managed IndexTTS-2 environments remain supported.
 IndexTTS 2.5 is not bundled because its source environment and model weights
 require substantial disk space.
 
-1. Open **Model Catalogue → Engines**.
+1. Open **Model Catalogue**.
 2. Expand **IndexTTS 2.5** and select **Install**.
 3. Keep VoiceStudio open while source, dependencies, and weights download.
 
@@ -97,7 +97,7 @@ IndexTTS-2 installations continue receiving their `target_tokens` control.
 
 ### Engine unavailable
 
-Use **Model Catalogue → Engines → IndexTTS 2.5 → Install**. For a manual install,
+Use **Model Catalogue → IndexTTS 2.5 → Install**. For a manual install,
 confirm that the configured directory contains:
 
 ```text

docs/engines/kittentts.md

@@ -20,7 +20,7 @@ only — but a much faster and much smaller install.
 pip install kittentts
 ```
 
-Then select the engine via **Model Catalogue → Engines** or
+Then select the engine via **Model Catalogue** or
 `OMNIVOICE_TTS_BACKEND=kittentts`.
 
 ## Voices

docs/engines/mlx-audio.md

@@ -19,7 +19,7 @@ platforms never reports as available
 pip install mlx-audio
 ```
 
-Then select the engine via **Model Catalogue → Engines** or
+Then select the engine via **Model Catalogue** or
 `OMNIVOICE_TTS_BACKEND=mlx-audio`.
 
 ## Model selection
@@ -36,7 +36,7 @@ One backend hosts many models. The curated set:
 | `melotts` | `mlx-community/MeloTTS-English-v3-MLX` | lightweight VITS |
 | `outetts` | `mlx-community/Llama-OuteTTS-1.0-1B-4bit` | LM-based |
 
-Pick a model in the **Model Catalogue → Engines** curated picker
+Pick a model in the **Model Catalogue** curated picker
 ([#981](https://github.com/debpalash/VoiceStudio/issues/981)) or set
 `OMNIVOICE_MLX_AUDIO_MODEL` to either a curated key (`kokoro`) or any full
 HF repo id. The env var overrides the persisted UI choice.

docs/engines/mlx-whisper.md

@@ -11,7 +11,7 @@ on every Apple Silicon machine
 ## Selecting it
 
 - Nothing to do on Apple Silicon — auto-detect prefers it there.
-- Or explicitly: **Model Catalogue → Engines**, ASR tab → **Use**, or
+- Or explicitly: **Model Catalogue**, ASR tab → **Use**, or
   `OMNIVOICE_ASR_BACKEND=mlx-whisper`.
 
 ## Best at

docs/engines/moonshine.md

@@ -11,7 +11,7 @@ where even int8 whisper-large is too slow.
 - Install one of the runtimes into the app venv:
   `uv pip install moonshine-onnx` (lighter, tried first) or
   `moonshine-voice`.
-- Then **Model Catalogue → Engines**, ASR tab → **Use** on the Moonshine row,
+- Then **Model Catalogue**, ASR tab → **Use** on the Moonshine row,
   or `OMNIVOICE_ASR_BACKEND=moonshine`.
 
 Auto-detect never picks it; it's an explicit opt-in.

docs/engines/moss-tts-nano.md

@@ -25,7 +25,7 @@ cd MOSS-TTS-Nano
 uv pip install -e .
 ```
 
-Then select the engine via **Model Catalogue → Engines** or
+Then select the engine via **Model Catalogue** or
 `OMNIVOICE_TTS_BACKEND=moss-tts-nano`.
 
 ## Model selection

docs/engines/moss-tts-v15.md

@@ -13,7 +13,7 @@ interpreter, so MOSS runs behind
 `backend/services/subprocess_backend.py::SubprocessBackend`.
 
 > **Opt-in, and never a default.** MOSS-TTS-v1.5 is selected explicitly in
-> **Model Catalogue → Engines** (or `OMNIVOICE_TTS_BACKEND=moss-tts-v15`). It is
+> **Model Catalogue** (or `OMNIVOICE_TTS_BACKEND=moss-tts-v15`). It is
 > not part of the default install and does not change VoiceStudio's
 > out-of-the-box behaviour on any platform.
 
@@ -75,7 +75,7 @@ into an isolated venv on demand.
    [Environment]::SetEnvironmentVariable("OMNIVOICE_MOSS_TTS_V15_DIR","$env:USERPROFILE\code\MOSS-TTS","User")
    ```
 
-5. Restart VoiceStudio. MOSS-TTS-v1.5 appears in **Model Catalogue → Engines** with
+5. Restart VoiceStudio. MOSS-TTS-v1.5 appears in **Model Catalogue** with
    `available: true` and `isolation_mode: subprocess`.
 
 ## Venv resolution order

docs/engines/nemo-parakeet.md

@@ -29,7 +29,7 @@ instead — same model family, no NeMo dependency:
 Only meaningful if you've set up `nemo_toolkit[asr]` in a **separate,
 dedicated Python environment** that runs the backend:
 
-- **Model Catalogue → Engines**, ASR tab → **Use** on the Parakeet TDT row, or
+- **Model Catalogue**, ASR tab → **Use** on the Parakeet TDT row, or
 - `OMNIVOICE_ASR_BACKEND=nemo-parakeet`.
 
 Auto-detect never picks it; it's an explicit opt-in.

docs/engines/omnivoice-gguf.md

@@ -32,10 +32,10 @@ against the same table (an F32 reference quant, ~3.2 GB, is override-only).
 ## Setup
 
 Nothing to install: installer and CI builds bundle the binary for your
-platform. Select the engine via **Model Catalogue → Engines** or
+platform. Select the engine via **Model Catalogue** or
 `OMNIVOICE_TTS_BACKEND=omnivoice-gguf`. The quant weights download on first
 use (see [downloading-models.md](../downloading-models.md)) — install them
-ahead of time from **Model Catalogue → Models** if you want the first
+ahead of time from **Model Catalogue → Downloaded weights** if you want the first
 generation to be quick; a long first render is the download, not a hang.
 
 **Source checkouts:** the repo ships zero-byte placeholders in `bin/` — real

docs/engines/omnivoice-subprocess.md

@@ -32,7 +32,7 @@ lower call overhead.
 
 ## Selecting it
 
-- **Model Catalogue → Engines** on CUDA, ROCm, or CPU, or
+- **Model Catalogue** on CUDA, ROCm, or CPU, or
 - `OMNIVOICE_TTS_BACKEND=omnivoice-subprocess`
 
 The explicit engine is opt-in on CUDA, ROCm, and CPU. On Apple Silicon it is

docs/engines/omnivoice.md

@@ -39,7 +39,7 @@ quantized native binary with a much smaller memory footprint.
 OmniVoice is the default, so normally there is nothing to do. If you switched
 away and want it back:
 
-- **Model Catalogue → Engines**, or
+- **Model Catalogue**, or
 - set `OMNIVOICE_TTS_BACKEND=omnivoice`.
 
 The env var overrides the persisted UI choice.
@@ -103,7 +103,7 @@ The env var overrides the persisted UI choice.
   above — switch to OmniVoice GGUF or close other GPU apps.
 - First generation is slow: the first call downloads multi-GB weights. To
   keep the first render quick, install the model ahead of time from
-  **Model Catalogue → Models** — a long first generate is almost always the
+  **Model Catalogue → Downloaded weights** — a long first generate is almost always the
   download, not a hang.
 - General install issues: [install/troubleshooting.md](../install/troubleshooting.md).
 

docs/engines/openai-compatible-asr.md

@@ -9,7 +9,7 @@ model installation and compute.
 
 ## Setup
 
-Everything lives on one screen — **Model Catalogue → Engines**, **ASR** tab:
+Everything lives on one screen — **Model Catalogue**, **ASR** tab:
 
 1. The **OpenAI-compatible (remote server)** row shows as unavailable until
    a server is configured. The config panel appears **below the engine

docs/engines/parakeet-mlx.md

@@ -12,10 +12,10 @@ Apple Silicon source installs since 0.3.22**.
 
 ## Selecting it
 
-- **Model Catalogue → Engines**, ASR tab → **Use** on the Parakeet TDT v3
+- **Model Catalogue**, ASR tab → **Use** on the Parakeet TDT v3
   (MLX) row, or `OMNIVOICE_ASR_BACKEND=parakeet-mlx`.
 - **Dictation prefers it automatically**: once the model weights are
-  installed (Model Catalogue → Models — the auto-pick never triggers a
+  installed (Model Catalogue → Downloaded weights — the auto-pick never triggers a
   download), live dictation/capture uses it whenever your system language is
   one of the 25 covered European languages. Other languages keep the
   multilingual Whisper engine, so dictation coverage never regresses.

docs/engines/pockettts.md

@@ -24,18 +24,18 @@ for this model.
    uv sync --extra pockettts
    ```
 
-   (Or enable it from **Model Catalogue → Engines**.)
+   (Or enable it from **Model Catalogue**.)
 
 2. **Accept the license in-app**
    ([#1306](https://github.com/debpalash/VoiceStudio/issues/1306)). The code
    is MIT and the weights are CC-BY-4.0, but the weights are **gated on
    HuggingFace** behind an access agreement with an acceptable-use clause.
    VoiceStudio surfaces this before first use: the engine stays unavailable
-   until you review and accept in **Model Catalogue → Engines → PocketTTS**.
+   until you review and accept in **Model Catalogue → PocketTTS**.
    You also need HuggingFace access to the gated repo (see
    [downloading-models.md](../downloading-models.md) for token setup).
 
-3. Select the engine via **Model Catalogue → Engines** or
+3. Select the engine via **Model Catalogue** or
    `OMNIVOICE_TTS_BACKEND=pockettts`.
 
 ## Platform notes
@@ -77,7 +77,7 @@ for this model.
 ## Troubleshooting
 
 - "pocket_tts package not installed": run the `uv sync` above.
-- "license not accepted": open **Model Catalogue → Engines → PocketTTS**
+- "license not accepted": open **Model Catalogue → PocketTTS**
   and review/accept.
 - Timeouts on a slow connection: raise
   `OMNIVOICE_POCKETTTS_RECV_TIMEOUT_S` for the first (download-heavy) run.

docs/engines/pytorch-whisper.md

@@ -9,7 +9,7 @@ genuinely uses **AMD ROCm** GPUs, so auto-detect picks it on ROCm hosts
 
 ## Selecting it
 
-- **Model Catalogue → Engines**, ASR tab → **Use** on the PyTorch Whisper
+- **Model Catalogue**, ASR tab → **Use** on the PyTorch Whisper
   row, or `OMNIVOICE_ASR_BACKEND=pytorch-whisper`.
 - `OMNIVOICE_ASR_BACKEND=omnivoice` is accepted as a compatibility alias and
   selects this same PyTorch-native ASR path on ROCm hosts.

docs/engines/sherpa-onnx-asr.md

@@ -10,8 +10,8 @@ partials either way.
 ## Selecting it
 
 - Ensure `sherpa-onnx` is installed (`uv add sherpa-onnx` on source installs).
-- Pick a dictation model in the app (Model Catalogue → Models lists the
-  selectable set below), or **Model Catalogue → Engines**, ASR tab → **Use**, or
+- Pick a dictation model in the app (Model Catalogue → Downloaded weights lists the
+  selectable set below), or **Model Catalogue**, ASR tab → **Use**, or
   pin `OMNIVOICE_ASR_BACKEND=sherpa-onnx-asr`.
 - `OMNIVOICE_SHERPA_ASR_MODEL` selects the model — default
   `sherpa-whisper-tiny`.

docs/engines/sherpa-onnx.md

@@ -30,7 +30,7 @@ TTS model directory.
    export OMNIVOICE_SHERPA_MODEL=/path/to/model-dir
    ```
 
-4. Select the engine via **Model Catalogue → Engines** or
+4. Select the engine via **Model Catalogue** or
    `OMNIVOICE_TTS_BACKEND=sherpa-onnx`.
 
 The directory must contain `model.onnx` and `tokens.txt`. Sherpa-ONNX ships

docs/engines/supertonic3.md

@@ -19,16 +19,16 @@ crashes and cold init never block the rest of VoiceStudio.
    uv sync --extra supertonic
    ```
 
-   (Or enable it from **Model Catalogue → Engines**, which installs the
+   (Or enable it from **Model Catalogue**, which installs the
    pinned `supertonic` wheel for you.)
 
 2. **Accept the license in-app.** First use is gated behind an explicit
    acceptance dialog: the inference SDK is MIT, but the model weights are
    **OpenRAIL-M**, which carries use restrictions. The engine stays
-   unavailable until you review and accept in **Model Catalogue → Engines →
+   unavailable until you review and accept in **Model Catalogue →
    Supertonic-3**.
 
-3. Select the engine via **Model Catalogue → Engines** or
+3. Select the engine via **Model Catalogue** or
    `OMNIVOICE_TTS_BACKEND=supertonic3`.
 
 The first synthesis cold-downloads ~400 MB of model weights, pinned to an
@@ -66,7 +66,7 @@ log line.
 
 - "supertonic package not installed": run the `uv sync` above or enable
   from the Model Catalogue.
-- "license not accepted": open **Model Catalogue → Engines → Supertonic-3**
+- "license not accepted": open **Model Catalogue → Supertonic-3**
   and accept.
 - Other issues: [install/troubleshooting.md](../install/troubleshooting.md).
 

docs/engines/voxcpm2.md

@@ -31,7 +31,7 @@ pip install "voxcpm>=2.0.3"
 
 That is a version **floor**, not a pin — an older install still works, but
 the engine logs an upgrade hint at load time. Then select the engine via
-**Model Catalogue → Engines** or `OMNIVOICE_TTS_BACKEND=voxcpm2`.
+**Model Catalogue** or `OMNIVOICE_TTS_BACKEND=voxcpm2`.
 
 ## Model selection
 

docs/engines/whisperx.md

@@ -8,7 +8,7 @@ prefers it wherever CTranslate2 can use the GPU.
 
 ## Selecting it
 
-- **Model Catalogue → Engines**, ASR tab → **Use** on the WhisperX row, or
+- **Model Catalogue**, ASR tab → **Use** on the WhisperX row, or
 - pin it with `OMNIVOICE_ASR_BACKEND=whisperx` (the env var always wins over
   the Settings pick; with neither set, auto-detect chooses per-hardware).
 

docs/expressive-speech.md

@@ -129,7 +129,7 @@ the default engine's taxonomy, so free-text instruct currently needs the API
 Setup: clone + install [CosyVoice](https://github.com/FunAudioLLM/CosyVoice)
 (non-trivial: `git clone --recursive`, its requirements, SoX), then set
 `OMNIVOICE_COSYVOICE_MODEL` to the model directory and select it in
-Model Catalogue → Engines. CUDA or CPU; MPS is unverified upstream.
+Model Catalogue. CUDA or CPU; MPS is unverified upstream.
 
 ### VoxCPM2 (opt-in)
 

docs/install/docker.md

@@ -215,7 +215,7 @@ ROCm container `omnivoice-studio-rocm` (CPU: `omnivoice-studio`, NVIDIA:
 name means torch can see the GPU.) That check alone isn't proof the app is
 using it: **Settings → Performance & Device** shows the device VoiceStudio
 actually resolved.
-**Model Catalogue → Engines** should report both `omnivoice` and
+**Model Catalogue** should report both `omnivoice` and
 `omnivoice-subprocess` as accelerated on ROCm, rather than a CPU-fallback
 warning.
 If it reads `cpu` while the command above prints `True`, the backend log line

docs/install/macos.md

@@ -157,7 +157,7 @@ without the quarantine step.
 - **Apple Silicon (M-series):** VoiceStudio automatically picks the `mlx-whisper`
   and `mlx-audio` backends where available — these use the Apple Neural Engine
   and Metal Performance Shaders for ~2× the throughput of the CPU path.
-  Installing the **Parakeet TDT v3 (MLX)** model from **Model Catalogue → Models**
+  Installing the **Parakeet TDT v3 (MLX)** model from **Model Catalogue → Downloaded weights**
   additionally makes dictation/capture prefer the `parakeet-mlx` engine
   (25 European languages, word timestamps, ~2 GB unified memory) — it is never
   downloaded without that explicit install, and it is only auto-preferred when
@@ -170,7 +170,7 @@ without the quarantine step.
   works only when pointed at a remote backend (**Settings → Sharing → Remote
   backend**).
 
-The picker in **Model Catalogue → Engines** shows which backend is active.
+The picker in **Model Catalogue** shows which backend is active.
 
 ## Hugging Face token (optional but recommended)
 

docs/install/troubleshooting.md

@@ -89,7 +89,7 @@ uv pip install --reinstall transformers
 ```
 
 Or, as a quick workaround, switch ASR to **faster-whisper** in
-**Model Catalogue → Models**. If it recurs, add the backend **`.venv`** to your
+**Model Catalogue → Downloaded weights**. If it recurs, add the backend **`.venv`** to your
 antivirus exclusions (see §1). Newer builds classify this error and show the
 reinstall hint directly instead of a bare path + "try restarting".
 
@@ -141,7 +141,7 @@ uv pip install --reinstall pytorch-lightning
 
 If it recurs, add the backend **`.venv`** to your antivirus exclusions (see
 §1). Since this fix landed the app also degrades gracefully: WhisperX is
-marked unavailable (Model Catalogue → Engines shows why, with this repair command)
+marked unavailable (Model Catalogue shows why, with this repair command)
 and dubbing automatically falls through to **faster-whisper** instead of
 failing outright.
 
@@ -205,7 +205,7 @@ itself does not grant access.
    license and prohibited-use conditions, share the requested contact details,
    and accept the conditions.
 2. Open **Settings → API Keys** and save a read token from that same account.
-3. Open **Model Catalogue → Engines**, review and accept the PocketTTS terms locally,
+3. Open **Model Catalogue**, review and accept the PocketTTS terms locally,
    then retry. VoiceStudio stores this acknowledgement only on your machine.
 
 ## 3. Gatekeeper quarantine on macOS
@@ -408,7 +408,7 @@ Intel-Mac wheels, so this entry only applies to historical installs (see
 ## 10. Windows: `Could not locate cudnn_ops_infer64_8.dll` during transcription
 
 **Symptom:** on Windows + NVIDIA, transcription/dubbing fails and the backend
-log shows `Could not locate cudnn_ops_infer64_8.dll`. Model Catalogue → Models shows
+log shows `Could not locate cudnn_ops_infer64_8.dll`. Model Catalogue → Downloaded weights shows
 WhisperX or faster-whisper selected.
 
 On builds before this was fixed, the failure looked much worse than a failed
@@ -444,7 +444,7 @@ uv pip install --no-deps --python .venv\Scripts\python.exe --target .venv\Lib\si
 (On Linux the target is `.venv/lib/pythonX.Y/site-packages/cudnn8_compat`.)
 
 Or sidestep cuDNN 8 entirely: switch the ASR backend to **PyTorch Whisper** in
-**Model Catalogue → Models**. It runs on PyTorch's own stack (cuDNN 9, bundled with
+**Model Catalogue → Downloaded weights**. It runs on PyTorch's own stack (cuDNN 9, bundled with
 torch) and needs no cuDNN-8 DLL — it loads its Whisper pipeline on demand (no
 extra env var).
 
@@ -600,12 +600,12 @@ did was `generate:start (audio)`, a dub, or a dictation.
 
 **Fix — reduce ASR load (any one of these):**
 
-1. **Pick a smaller ASR model / engine** in **Model Catalogue → Models** — e.g.
+1. **Pick a smaller ASR model / engine** in **Model Catalogue → Downloaded weights** — e.g.
    faster-whisper **medium** or **small**, instead of large-v3. Biggest win on
    low-VRAM GPUs.
 2. **Free VRAM**: **Flush the TTS model** before dubbing so ASR isn't competing
    for memory (top toolbar → Flush → "Unload all + flush", or per-model from
-   Model Catalogue → Models — see [Flush caches / Unload resident model](../performance.md#flush-caches--unload-resident-model)
+   Model Catalogue → Downloaded weights — see [Flush caches / Unload resident model](../performance.md#flush-caches--unload-resident-model)
    for exactly what it frees and the API equivalents for scripts), or
 3. **Run ASR on CPU** (slower but reliable) if your GPU is small.
 4. **Test with a 10-second clip** first — if that returns quickly, it confirms a
@@ -675,7 +675,7 @@ hammering — on a 1-worker machine, concurrent requests serialize by design.
 recovering the underlying hang — the wedged thread keeps its VRAM until the app
 exits. The error message will then recommend switching the ASR engine to
 **Faster-Whisper (crash-isolated subprocess)** (`faster-whisper-isolated`) in
-**Model Catalogue → Engines**: it runs transcription in a separate process that can be
+**Model Catalogue**: it runs transcription in a separate process that can be
 force-killed to reclaim a hung transcribe *and* its VRAM, at a small per-call
 overhead. It reuses your existing faster-whisper install (nothing extra to
 download). VoiceStudio never switches engines automatically — this stays your

docs/migration/real-time-voice-cloning.md

@@ -25,7 +25,7 @@ on your own machine: no accounts, no API keys, no cloud.
 |---|---|---|
 | Speaker encoder + reference utterance | Reference clip in the **Voice Clone** workflow ("From audio") | No separate embedding step — zero-shot engines condition on the clip directly |
 | Saved speaker embeddings (`.npy`) | **Voice Profiles** — save a clone once, reuse it everywhere | Exportable as portable `.ovsvoice` bundles |
-| Synthesizer + vocoder choice (Tacotron 2 · WaveRNN / Griffin-Lim) | **TTS engine choice** — Model Catalogue → Engines | 14 engines, from CPU-realtime to GPU heavyweights; per-engine GPU preflight |
+| Synthesizer + vocoder choice (Tacotron 2 · WaveRNN / Griffin-Lim) | **TTS engine choice** — Model Catalogue | 14 engines, from CPU-realtime to GPU heavyweights; per-engine GPU preflight |
 | The Toolbox GUI (`demo_toolbox.py`) | The app itself | Record or drop a clip, type text, synthesize — same loop, no `python demo_toolbox.py` |
 | `demo_cli.py` / scripting your own pipeline | Local REST API (OpenAI-compatible, `http://localhost:3900/v1`), `omnivoice-infer` CLI, MCP server | See the [API section of the README](../../README.md#openai-api) |
 | Training your own encoder / synthesizer / vocoder | Partial — see ["What RTVC did that VoiceStudio doesn't"](#what-rtvc-did-that-omnivoice-doesnt) | Fine-tuning the bundled model is documented; RTVC-style three-stage research training is not what this project is |

docs/performance.md

@@ -38,7 +38,7 @@ Before touching any knob, check these — they account for most slowness reports
    - **Settings → About → Run self-check** (the `/system/diagnose` endpoint)
      warns explicitly: *"cpu (no GPU acceleration detected)"* with a hint
      about drivers.
-   - **Model Catalogue → Engines** shows a routing badge per engine — "GPU active",
+   - **Model Catalogue** shows a routing badge per engine — "GPU active",
      "CPU fallback", or "CPU" — with the *reason* shown as small text under
      the badge (full text on hover).
    Note: **GPU acceleration on Windows is NVIDIA/CUDA-only** — AMD and Intel
@@ -183,7 +183,7 @@ drain, or restart the backend, and then Flush.
   - **Unload all + flush** — the above **plus** fully unloads the resident
     TTS model. Frees the most memory; the next generation pays the ~8 s
     reload.
-- **Model Catalogue → Models** — rows whose weights are resident right now show an
+- **Model Catalogue → Downloaded weights** — rows whose weights are resident right now show an
   "In memory" badge with the same per-model **Unload** button.
 
 **From a script** (the local API on port 3900), the same operations:
@@ -234,7 +234,7 @@ CPU ASR, the crash-isolated ASR engine).
   to 3-4 concurrent generations (API/batch workloads); ≤10 GB deliberately
   serializes.
 - **CPU-only**: expect ~2x slower than MPS, more against CUDA. Prefer the
-  smaller/faster engines (see Model Catalogue → Engines) and short reference clips.
+  smaller/faster engines (see Model Catalogue) and short reference clips.
 
 ## Measuring instead of guessing
 

frontend/src/components/EngineQuickSwitch.jsx

@@ -218,7 +218,7 @@ export default function EngineQuickSwitch({
               type="button"
               onClick={() => {
                 setOpen(false);
-                useAppStore.getState().openCatalogue({ pane: 'engines', family });
+                useAppStore.getState().openCatalogue({ family });
               }}
               className="mt-[3px] flex items-center gap-[3px] border-0 bg-transparent px-[7px] py-[5px] text-left text-[11px] text-[color:var(--chrome-fg-muted)] hover:text-[color:var(--chrome-fg)]"
             >

frontend/src/components/Header.jsx

@@ -517,9 +517,7 @@ export default function Header({
                         className="flex items-center gap-1 border-0 bg-transparent p-2 text-xs text-[var(--chrome-fg-muted)] cursor-pointer hover:text-[var(--chrome-fg)]"
                         onClick={() => {
                           setFlushOpen(false);
-                          useAppStore
-                            .getState()
-                            .openCatalogue({ pane: 'engines', family: engineFamily });
+                          useAppStore.getState().openCatalogue({ family: engineFamily });
                         }}
                       >
                         {t('header.label_catalogue')} <ChevronRight size={12} />

frontend/src/components/catalogue/SetupSummary.jsx

@@ -0,0 +1,236 @@
+import React, { useEffect, useMemo, useState } from 'react';
+import { useQuery } from '@tanstack/react-query';
+import { useTranslation } from 'react-i18next';
+import { toast } from 'react-hot-toast';
+import { Download, RefreshCw } from 'lucide-react';
+import { apiJson } from '../../api/client';
+import { useEngines, useInstallModel, useRecommendations } from '../../api/hooks';
+import { useAppStore } from '../../store';
+import { Badge, Button } from '../../ui';
+
+/**
+ * SetupSummary — the Model Catalogue's first screen: what the app will use
+ * right now for speech, transcription, dictation and the language model, one
+ * line each, and a "Change" that jumps to that family's engine list below.
+ *
+ * Most users take the recommended engines and never come back; this is the
+ * view built for them. Everything that used to compete for attention on the
+ * old two-pane page (engine matrix, model store, stats) now sits under it,
+ * reached only through Change.
+ */
+const DEVICE_LABEL = { cuda: 'CUDA', mps: 'MPS', rocm: 'ROCm', cpu: 'CPU', directml: 'DirectML' };
+
+const STATUS_TONE = {
+  ready: 'success',
+  cpu: 'warn',
+  setup: 'warn',
+  off: 'neutral',
+  missing: 'warn',
+};
+
+/** Pure: reduce one family's /engines payload to {name, device, status}. */
+export function summarizeFamily(family, familyData) {
+  const active = familyData?.active;
+  const backends = familyData?.backends || [];
+  if (family === 'llm' && (!active || active === 'off')) {
+    return { name: null, device: null, status: 'off' };
+  }
+  const entry = backends.find((b) => b.id === active);
+  const name = entry?.display_name || active || null;
+  if (!entry || entry.available === false) return { name, device: null, status: 'setup' };
+  const routing = entry.routing_status;
+  const device =
+    routing === 'n/a' ? 'remote' : DEVICE_LABEL[entry.effective_device] || entry.effective_device;
+  return {
+    name,
+    device: device || null,
+    status: routing === 'cpu_fallback' ? 'cpu' : 'ready',
+  };
+}
+
+/** Pure: the dictation line from /dictation/models + the model_id pref. */
+export function summarizeDictation(data, modelId) {
+  const models = Array.isArray(data?.models) ? data.models : [];
+  const selected =
+    models.find((m) => m.id === modelId) || models.find((m) => m.recommended) || models[0];
+  if (!selected) return { name: null, device: null, status: 'setup' };
+  if (data?.engine_available === false)
+    return { name: selected.label, device: null, status: 'setup' };
+  return {
+    name: selected.label,
+    device: DEVICE_LABEL.cpu,
+    status: selected.installed ? 'ready' : 'missing',
+  };
+}
+
+export default function SetupSummary({ onChange }) {
+  const { t } = useTranslation();
+  const enginesQuery = useEngines();
+  const recoQuery = useRecommendations();
+  const installMutation = useInstallModel();
+  const [installing, setInstalling] = useState(false);
+
+  const modelId = useAppStore((s) => s.dictationModelId);
+  const loadPrefs = useAppStore((s) => s.loadDictationPrefs);
+  const dictationLoaded = useAppStore((s) => s.dictationLoaded);
+  useEffect(() => {
+    if (!dictationLoaded) loadPrefs?.();
+  }, [dictationLoaded, loadPrefs]);
+  // Same query key as the quick-menu picker so both read one cache entry.
+  const dictationQuery = useQuery({
+    queryKey: ['dictation-models'],
+    queryFn: () => apiJson('/dictation/models'),
+    staleTime: 10_000,
+    retry: false,
+  });
+
+  const engines = enginesQuery.data;
+  const rows = useMemo(
+    () => [
+      {
+        key: 'speech',
+        family: 'tts',
+        label: t('header.speech'),
+        ...summarizeFamily('tts', engines?.tts),
+      },
+      {
+        key: 'transcription',
+        family: 'asr',
+        label: t('projects.transcription'),
+        ...summarizeFamily('asr', engines?.asr),
+      },
+      {
+        key: 'dictation',
+        family: 'asr',
+        label: t('settings.dictation'),
+        ...summarizeDictation(dictationQuery.data, modelId),
+      },
+      {
+        key: 'llm',
+        family: 'llm',
+        label: t('models.role_llm'),
+        ...summarizeFamily('llm', engines?.llm),
+      },
+    ],
+    [t, engines, dictationQuery.data, modelId],
+  );
+
+  const statusLabel = {
+    ready: t('models.ready_badge'),
+    cpu: t('catalogue.status_cpu'),
+    setup: t('catalogue.status_setup'),
+    off: t('catalogue.status_off'),
+    missing: t('models.not_installed'),
+  };
+
+  const reco = recoQuery.data;
+  const missing = reco?.models?.filter((m) => !m.installed) || [];
+  const installRest = async () => {
+    if (missing.length === 0) return;
+    setInstalling(true);
+    try {
+      await Promise.all(missing.map((m) => installMutation.mutateAsync(m.repo_id)));
+      toast.success(t('models.started_downloading', { count: missing.length }));
+    } catch (e) {
+      toast.error(t('models.install_failed', { message: e?.message || e }));
+    } finally {
+      setInstalling(false);
+    }
+  };
+
+  return (
+    <section
+      aria-labelledby="setup-summary-title"
+      data-testid="setup-summary"
+      className="mb-[32px] rounded-[14px] border border-border bg-card/40 px-[20px] py-[6px] font-sans"
+    >
+      <h2 id="setup-summary-title" className="sr-only">
+        {t('catalogue.summary_title')}
+      </h2>
+      <dl className="m-0 divide-y divide-border">
+        {rows.map((row) => (
+          <div
+            key={row.key}
+            data-testid={`setup-row-${row.key}`}
+            className="flex flex-wrap items-center gap-x-[16px] gap-y-[4px] py-[10px]"
+          >
+            <dt className="w-[128px] shrink-0 text-sm font-semibold text-foreground">
+              {row.label}
+            </dt>
+            <dd className="m-0 flex min-w-0 flex-1 flex-wrap items-center gap-x-[8px] text-sm text-muted-foreground">
+              {enginesQuery.isLoading && !engines && row.key !== 'dictation' ? (
+                <span>{t('common.loading')}</span>
+              ) : (
+                <>
+                  {row.name && (
+                    <span
+                      className="truncate text-foreground"
+                      data-testid={`setup-name-${row.key}`}
+                    >
+                      {row.name}
+                    </span>
+                  )}
+                  {row.device && (
+                    <>
+                      {row.name && <span aria-hidden="true">·</span>}
+                      <span className="font-mono text-[11px] uppercase tracking-[0.03em]">
+                        {row.device === 'remote' ? t('catalogue.device_remote') : row.device}
+                      </span>
+                    </>
+                  )}
+                  {(row.name || row.device) && <span aria-hidden="true">·</span>}
+                  <Badge
+                    tone={STATUS_TONE[row.status]}
+                    size="xs"
+                    data-testid={`setup-status-${row.key}`}
+                  >
+                    {statusLabel[row.status]}
+                  </Badge>
+                </>
+              )}
+            </dd>
+            <Button
+              variant="subtle"
+              size="sm"
+              onClick={() => onChange?.(row.family)}
+              data-testid={`setup-change-${row.key}`}
+            >
+              {t('catalogue.change')}
+            </Button>
+          </div>
+        ))}
+      </dl>
+      {reco && (
+        <div
+          data-testid="setup-reco"
+          className="flex flex-wrap items-center gap-x-[12px] gap-y-[6px] border-t border-border py-[10px] text-xs text-muted-foreground"
+        >
+          <span className="min-w-0 flex-1">
+            {reco.all_installed || missing.length === 0
+              ? t('models.reco_installed_for', { device: reco.device?.label })
+              : t('catalogue.reco_line', {
+                  installed: reco.models.length - missing.length,
+                  total: reco.models.length,
+                  device: reco.device?.label,
+                  remaining: reco.download_gb_remaining,
+                })}
+          </span>
+          {missing.length > 0 && (
+            <Button
+              variant="primary"
+              size="sm"
+              onClick={installRest}
+              disabled={installing}
+              leading={
+                installing ? <RefreshCw size={11} className="spinner" /> : <Download size={11} />
+              }
+              data-testid="setup-install-rest"
+            >
+              {installing ? t('models.starting') : t('catalogue.install_rest')}
+            </Button>
+          )}
+        </div>
+      )}
+    </section>
+  );
+}

frontend/src/components/catalogue/SetupSummary.test.jsx

@@ -0,0 +1,208 @@
+import React from 'react';
+import { describe, it, expect, vi, beforeEach } from 'vitest';
+import { render, screen, fireEvent, waitFor, within } from '@testing-library/react';
+import { QueryClient, QueryClientProvider } from '@tanstack/react-query';
+import { I18nextProvider } from 'react-i18next';
+import i18n from '../../i18n';
+
+vi.mock('react-hot-toast', () => ({ toast: { success: vi.fn(), error: vi.fn() } }));
+
+let enginesData = null;
+let recoData = null;
+const installMutate = vi.fn().mockResolvedValue({});
+vi.mock('../../api/hooks', () => ({
+  useEngines: () => ({ data: enginesData, isLoading: false }),
+  useRecommendations: () => ({ data: recoData }),
+  useInstallModel: () => ({ mutateAsync: installMutate }),
+}));
+vi.mock('../../api/client', () => ({ apiJson: vi.fn() }));
+
+import { apiJson } from '../../api/client';
+import { useAppStore } from '../../store';
+import SetupSummary, { summarizeDictation, summarizeFamily } from './SetupSummary';
+
+const t = i18n.t.bind(i18n);
+
+const ENGINES = {
+  tts: {
+    active: 'indextts',
+    backends: [
+      {
+        id: 'indextts',
+        display_name: 'IndexTTS 2',
+        available: true,
+        effective_device: 'mps',
+        routing_status: 'accelerated',
+      },
+    ],
+  },
+  asr: {
+    active: 'faster-whisper',
+    backends: [
+      {
+        id: 'faster-whisper',
+        display_name: 'Faster-Whisper',
+        available: true,
+        effective_device: 'cpu',
+        routing_status: 'cpu_fallback',
+      },
+    ],
+  },
+  llm: { active: 'off', backends: [] },
+};
+
+const DICTATION = {
+  engine_available: true,
+  models: [
+    { id: 'sherpa-whisper-tiny', label: 'Whisper Tiny', installed: true, recommended: true },
+    { id: 'sherpa-parakeet-tdt-v3', label: 'Parakeet TDT v3', installed: false },
+  ],
+};
+
+function mount(props = {}) {
+  const client = new QueryClient({ defaultOptions: { queries: { retry: false } } });
+  return render(
+    <QueryClientProvider client={client}>
+      <I18nextProvider i18n={i18n}>
+        <SetupSummary onChange={vi.fn()} {...props} />
+      </I18nextProvider>
+    </QueryClientProvider>,
+  );
+}
+
+describe('SetupSummary', () => {
+  beforeEach(() => {
+    vi.clearAllMocks();
+    enginesData = ENGINES;
+    recoData = null;
+    apiJson.mockResolvedValue(DICTATION);
+    useAppStore.setState({ dictationModelId: 'sherpa-whisper-tiny', dictationLoaded: true });
+  });
+
+  it('shows one line per capability: engine, device, and a plain status word', async () => {
+    mount();
+    const speech = screen.getByTestId('setup-row-speech');
+    expect(within(speech).getByTestId('setup-name-speech')).toHaveTextContent('IndexTTS 2');
+    expect(speech).toHaveTextContent('MPS');
+    expect(within(speech).getByTestId('setup-status-speech')).toHaveTextContent(
+      t('models.ready_badge'),
+    );
+
+    const asr = screen.getByTestId('setup-row-transcription');
+    expect(asr).toHaveTextContent('Faster-Whisper');
+    expect(within(asr).getByTestId('setup-status-transcription')).toHaveTextContent(
+      t('catalogue.status_cpu'),
+    );
+
+    const llm = screen.getByTestId('setup-row-llm');
+    expect(within(llm).getByTestId('setup-status-llm')).toHaveTextContent(
+      t('catalogue.status_off'),
+    );
+    expect(within(llm).queryByTestId('setup-name-llm')).toBeNull();
+
+    await waitFor(() =>
+      expect(screen.getByTestId('setup-name-dictation')).toHaveTextContent('Whisper Tiny'),
+    );
+    expect(screen.getByTestId('setup-status-dictation')).toHaveTextContent(t('models.ready_badge'));
+  });
+
+  it('flags an unavailable active engine and an undownloaded dictation model', async () => {
+    enginesData = {
+      ...ENGINES,
+      tts: { active: 'gone', backends: [{ id: 'gone', display_name: 'Gone', available: false }] },
+    };
+    useAppStore.setState({ dictationModelId: 'sherpa-parakeet-tdt-v3' });
+    mount();
+    expect(screen.getByTestId('setup-status-speech')).toHaveTextContent(
+      t('catalogue.status_setup'),
+    );
+    await waitFor(() =>
+      expect(screen.getByTestId('setup-status-dictation')).toHaveTextContent(
+        t('models.not_installed'),
+      ),
+    );
+  });
+
+  it('Change hands the row’s engine family to the page (dictation lives under ASR)', () => {
+    const onChange = vi.fn();
+    mount({ onChange });
+    for (const [key, family] of [
+      ['speech', 'tts'],
+      ['transcription', 'asr'],
+      ['dictation', 'asr'],
+      ['llm', 'llm'],
+    ]) {
+      fireEvent.click(screen.getByTestId(`setup-change-${key}`));
+      expect(onChange).toHaveBeenLastCalledWith(family);
+    }
+  });
+
+  it('summarises the recommended preset in one line and installs the rest in one click', async () => {
+    recoData = {
+      device: { label: 'Apple M2 Pro' },
+      all_installed: false,
+      download_gb_remaining: 2.4,
+      models: [
+        { repo_id: 'a/have', installed: true, size_gb: 1 },
+        { repo_id: 'b/need', installed: false, size_gb: 2.4 },
+      ],
+    };
+    mount();
+    const line = screen.getByTestId('setup-reco');
+    expect(line).toHaveTextContent(
+      t('catalogue.reco_line', { installed: 1, total: 2, device: 'Apple M2 Pro', remaining: 2.4 }),
+    );
+    fireEvent.click(screen.getByTestId('setup-install-rest'));
+    await waitFor(() => expect(installMutate).toHaveBeenCalledWith('b/need'));
+    expect(installMutate).toHaveBeenCalledTimes(1);
+  });
+
+  it('drops the install button once the preset is complete', () => {
+    recoData = {
+      device: { label: 'Apple M2 Pro' },
+      all_installed: true,
+      total_gb: 3.4,
+      models: [{ repo_id: 'a/have', installed: true, size_gb: 1 }],
+    };
+    mount();
+    expect(screen.getByTestId('setup-reco')).toHaveTextContent(
+      t('models.reco_installed_for', { device: 'Apple M2 Pro' }),
+    );
+    expect(screen.queryByTestId('setup-install-rest')).toBeNull();
+  });
+});
+
+describe('summarizeFamily / summarizeDictation', () => {
+  it('maps routing and availability to the four status words', () => {
+    expect(summarizeFamily('tts', ENGINES.tts)).toEqual({
+      name: 'IndexTTS 2',
+      device: 'MPS',
+      status: 'ready',
+    });
+    expect(summarizeFamily('asr', ENGINES.asr).status).toBe('cpu');
+    expect(summarizeFamily('llm', ENGINES.llm)).toEqual({
+      name: null,
+      device: null,
+      status: 'off',
+    });
+    expect(summarizeFamily('tts', { active: 'x', backends: [] })).toEqual({
+      name: 'x',
+      device: null,
+      status: 'setup',
+    });
+    expect(
+      summarizeFamily('asr', {
+        active: 'openai-compat',
+        backends: [{ id: 'openai-compat', available: true, routing_status: 'n/a' }],
+      }).device,
+    ).toBe('remote');
+  });
+
+  it('falls back to the recommended dictation model and reports engine trouble', () => {
+    expect(summarizeDictation(DICTATION, 'unknown').name).toBe('Whisper Tiny');
+    expect(summarizeDictation({ ...DICTATION, engine_available: false }, null).status).toBe(
+      'setup',
+    );
+    expect(summarizeDictation(null, null).status).toBe('setup');
+  });
+});

frontend/src/components/settings/CataloguePointer.jsx

@@ -1,49 +0,0 @@
-import React from 'react';
-import { Boxes } from 'lucide-react';
-import { useTranslation } from 'react-i18next';
-import { useAppStore } from '../../store';
-import { Button } from '../../ui';
-import { SETTINGS_SECTION_SURFACE } from './primitives';
-
-/**
- * CataloguePointer — the Settings-side signpost for what moved to the Model
- * Catalogue workspace.
- *
- * Engine selection and the model store are now a top-level workspace
- * (`pages/ModelCatalogue.jsx`), not Settings categories. The categories stay in
- * the Settings IA on purpose: "Engines" / "Models" is where users (and every
- * doc written before the move) look first, and search still has to find them.
- * They render this instead of the panels, so the trip is one click, not a hunt.
- *
- * @param {'engines'|'models'} area  which pane to open
- */
-export default function CataloguePointer({ area = 'engines' }) {
-  const { t } = useTranslation();
-  const openCatalogue = useAppStore((s) => s.openCatalogue);
-
-  return (
-    <section
-      className={`${SETTINGS_SECTION_SURFACE} flex flex-wrap items-center gap-[var(--space-4)]`}
-      data-slot="settings-section"
-      data-testid={`catalogue-pointer-${area}`}
-    >
-      <span
-        className="inline-flex h-[36px] w-[36px] shrink-0 items-center justify-center rounded-[var(--chrome-radius-pill)] bg-[color-mix(in_srgb,var(--chrome-accent)_12%,var(--chrome-bg))] text-[color:var(--chrome-accent)]"
-        aria-hidden="true"
-      >
-        <Boxes size={18} />
-      </span>
-      <div className="min-w-0 flex-auto">
-        <h3 className="m-0 [font-family:var(--font-sans)] text-[length:var(--text-md)] font-semibold text-[color:var(--chrome-fg)]">
-          {t('catalogue.moved_title')}
-        </h3>
-        <p className="m-0 mt-[2px] [font-family:var(--font-sans)] text-[length:var(--text-sm)] text-[color:var(--chrome-fg-muted)]">
-          {t(area === 'models' ? 'catalogue.moved_models_body' : 'catalogue.moved_engines_body')}
-        </p>
-      </div>
-      <Button variant="primary" onClick={() => openCatalogue(area)}>
-        {t(area === 'models' ? 'catalogue.open_models' : 'catalogue.open_engines')}
-      </Button>
-    </section>
-  );
-}

frontend/src/components/settings/CataloguePointer.test.jsx

@@ -1,28 +0,0 @@
-import { describe, it, expect, beforeEach } from 'vitest';
-import React from 'react';
-import { render, screen, fireEvent, act } from '@testing-library/react';
-
-import { useAppStore } from '../../store';
-import CataloguePointer from './CataloguePointer';
-
-describe('CataloguePointer', () => {
-  beforeEach(() => {
-    act(() => {
-      useAppStore.setState({ mode: 'settings', pendingCatalogueTab: null });
-    });
-  });
-
-  it('sends the Engines category to the catalogue engines pane', () => {
-    render(<CataloguePointer area="engines" />);
-    fireEvent.click(screen.getByRole('button', { name: 'Open Engines' }));
-    expect(useAppStore.getState().mode).toBe('catalogue');
-    expect(useAppStore.getState().pendingCatalogueTab).toBe('engines');
-  });
-
-  it('sends the Models category to the catalogue models pane', () => {
-    render(<CataloguePointer area="models" />);
-    fireEvent.click(screen.getByRole('button', { name: 'Open Models' }));
-    expect(useAppStore.getState().mode).toBe('catalogue');
-    expect(useAppStore.getState().pendingCatalogueTab).toBe('models');
-  });
-});

frontend/src/components/settings/LLMProvidersPanel.jsx

@@ -236,7 +236,7 @@ export default function LLMProvidersPanel() {
         <button
           type="button"
           className="cursor-pointer border-0 bg-transparent p-0 text-[length:var(--text-xs)] text-[color:var(--chrome-accent)] underline-offset-2 hover:underline"
-          onClick={() => useAppStore.getState().openCatalogue?.({ pane: 'engines', family: 'llm' })}
+          onClick={() => useAppStore.getState().openCatalogue?.({ family: 'llm' })}
           data-testid="llmp-open-catalogue"
         >
           {t('settings.llmp_catalogue_link')}

frontend/src/components/settings/ModelStoreTab.jsx

@@ -1,15 +1,13 @@
 import React, { useEffect, useState, useCallback, useMemo } from 'react';
-import { Cpu, RefreshCw, KeyRound } from 'lucide-react';
+import { RefreshCw, Search, Zap } from 'lucide-react';
 import { toast } from 'react-hot-toast';
 import { useTranslation } from 'react-i18next';
-import { openExternal } from '../../api/external';
 import { setupDownloadStreamUrl } from '../../api/setup';
 import { listLoadedModels, unloadLoadedModel } from '../../api/system';
 import { useModels, useRecommendations, useInstallModel, useDeleteModel } from '../../api/hooks';
 import { Button } from '../../ui';
-import { SettingsSection, SettingsInput, SETTINGS_SECTION_SURFACE } from './primitives';
+import { Input } from '@/components/ui/input';
 import { askConfirm } from './native';
-import { fmtBytes } from './models/format';
 import { computeRowRuntime } from './models/runtime';
 import {
   downloadKey,
@@ -18,19 +16,23 @@ import {
   isAutoPurgeTerminal,
 } from './models/downloadReducer';
 import { makeModelColumns } from './models/columns';
-import { groupModels } from './models/sections';
+import { FAMILY_SECTIONS, groupModels, modelSectionKey, scopeReco } from './models/sections';
 import RecoBanner from './models/RecoBanner';
 import ModelSection from './models/ModelSection';
-import VoicePreviewsPanel from './VoicePreviewsPanel';
 
 /**
- * Model store — every known HF model, grouped by capability (TTS / ASR /
- * Dictation / Diarisation), with install state and install / reinstall /
- * delete per row. The curated "for your system" preset leads (RecoBanner);
- * platform-incompatible rows collapse behind a per-section toggle. Per-model
- * download progress is pulled from the shared /setup/download-stream SSE.
+ * Model store — the downloadable weights for one engine family (or all of
+ * them when `family` is null), grouped by capability (TTS / ASR offline /
+ * Dictation streaming / Diarisation), with install state and install /
+ * reinstall / delete per row. The curated "for your system" preset leads
+ * (RecoBanner, scoped to the family); platform-incompatible rows collapse
+ * behind a per-section toggle. Per-model download progress is pulled from
+ * the shared /setup/download-stream SSE.
+ *
+ * Storage stats, the HF token and the voice-preview toggle used to sit on
+ * this list; they live in Settings → Storage / Credentials now.
  */
-export default function ModelStoreTab({ info, modelBadge, catalogueLayout = false }) {
+export default function ModelStoreTab({ info, family = null }) {
   const { t } = useTranslation();
   // Role labels — localized (diarization is an on-disk spelling alias for
   // diarisation; both map to the same label).
@@ -51,6 +53,7 @@ export default function ModelStoreTab({ info, modelBadge, catalogueLayout = fals
   const data = modelsQuery.data;
   const loading = modelsQuery.isLoading;
   const reco = recoQuery.data;
+  const familyReco = useMemo(() => scopeReco(reco, family), [reco, family]);
   const installMutation = useInstallModel();
   const deleteMutation = useDeleteModel();
 
@@ -83,36 +86,6 @@ export default function ModelStoreTab({ info, modelBadge, catalogueLayout = fals
     return () => clearInterval(iv);
   }, [hasActive]);
 
-  // HF token inline — compact input in the toolbar
-  const [hfToken, setHfToken] = useState('');
-  const [hfSaved, setHfSaved] = useState(false);
-  const [hfSaving, setHfSaving] = useState(false);
-  const [hfExpanded, setHfExpanded] = useState(false);
-  const saveHfToken = async () => {
-    const value = hfToken.trim();
-    if (!value) return;
-    setHfSaving(true);
-    try {
-      // One canonical token path: persist to the encrypted app store — the SAME
-      // store Settings → Credentials writes AND clears (/api/settings/hf-token).
-      // This toolbar used to POST /system/set-env (env var + HF-CLI file), a
-      // *second* store the Credentials "Clear" button couldn't reach, so a
-      // toolbar-set token silently outlived a Clear (a support-ticket generator).
-      // Now both entry points share one store with one clear path.
-      const { apiPost } = await import('../../api/client');
-      await apiPost('/api/settings/hf-token', { token: value });
-      toast.success(t('models.hf_token_set_toast'));
-      setHfSaved(true);
-      setHfToken('');
-      setHfExpanded(false);
-    } catch (e) {
-      toast.error(t('settings.save_failed', { message: e.message }));
-    } finally {
-      setHfSaving(false);
-    }
-  };
-  const hfTokenSet = hfSaved || info?.has_hf_token;
-
   // Open the progress stream once when the tab mounts; close on unmount.
   useEffect(() => {
     const es = new EventSource(setupDownloadStreamUrl());
@@ -297,8 +270,8 @@ export default function ModelStoreTab({ info, modelBadge, catalogueLayout = fals
   );
 
   const onInstallRecommended = async () => {
-    if (!reco) return;
-    const missing = reco.models.filter((m) => !m.installed);
+    if (!familyReco) return;
+    const missing = familyReco.models.filter((m) => !m.installed);
     if (missing.length === 0) {
       toast.success(t('models.recommended_installed'));
       return;
@@ -316,7 +289,12 @@ export default function ModelStoreTab({ info, modelBadge, catalogueLayout = fals
     }
   };
 
-  const allModels = React.useMemo(() => data?.models || [], [data]);
+  // The family's slice of the catalog (all of it when unscoped).
+  const allModels = React.useMemo(() => {
+    const models = data?.models || [];
+    const keep = family ? FAMILY_SECTIONS[family] || [] : null;
+    return keep ? models.filter((m) => keep.includes(modelSectionKey(m))) : models;
+  }, [data, family]);
   // Grouped catalog: TTS / ASR (offline transcription) / Dictation (streaming)
   // / Diarisation, with the search query applied per-section (pure helper —
   // matches the same fields the old global filter did).
@@ -367,202 +345,98 @@ export default function ModelStoreTab({ info, modelBadge, catalogueLayout = fals
   );
 
   if (loading && !data) {
-    if (catalogueLayout) {
-      return (
-        <div className="px-[2px] py-[24px] font-sans text-[var(--text-md)] text-[var(--chrome-fg-dim)]">
-          {t('common.loading')}
-        </div>
-      );
-    }
     return (
-      <SettingsSection icon={Cpu} title={t('settings.models')}>
-        <div className="settings-muted font-sans text-[var(--text-md)] text-[var(--chrome-fg-dim)]">
-          {t('common.loading')}
-        </div>
-      </SettingsSection>
+      <div className="px-[2px] py-[24px] font-sans text-sm text-muted-foreground">
+        {t('common.loading')}
+      </div>
     );
   }
   if (!data) return null;
 
   return (
-    <>
-      <section
-        className={
-          catalogueLayout
-            ? 'flex min-h-0 flex-col'
-            : `${SETTINGS_SECTION_SURFACE} flex min-h-[95%] flex-col`
-        }
-        data-slot="settings-section"
-        data-testid="model-list-panel"
-      >
-        <div
-          className={`flex flex-wrap items-center justify-between gap-[var(--space-3)] font-[family-name:var(--chrome-font-mono)] text-[length:var(--text-xs)] text-[var(--chrome-fg-muted)] max-[580px]:flex-col max-[580px]:items-start ${
-            catalogueLayout
-              ? 'mb-[24px] border-b border-[color-mix(in_srgb,var(--chrome-fg)_8%,transparent)] px-[2px] pb-[18px]'
-              : 'px-[2px] pb-[6px] pt-[2px]'
-          }`}
-        >
-          <div className="inline-flex flex-wrap items-center gap-[var(--space-2)]">
-            <span>
-              <strong className="font-semibold text-[var(--chrome-fg)]">
-                {fmtBytes(data.total_installed_bytes)}
-              </strong>
-            </span>
-            {data.disk_free_gb != null && (
-              <>
-                <span className="text-[var(--chrome-fg-dim)]">·</span>
-                <span title={t('models.disk_free_title')}>
-                  {t('models.disk_free', { size: `${data.disk_free_gb} GB` })}
-                </span>
-              </>
-            )}
-            <span className="text-[var(--chrome-fg-dim)]">·</span>
-            <span title={data.hf_cache_dir}>
-              <code className="font-[family-name:var(--chrome-font-mono)] text-[length:var(--text-xs)] text-[var(--chrome-fg)]">
-                {data.hf_cache_dir?.replace(/^\/Users\/[^/]+/, '~')}
-              </code>
-            </span>
-            {info && <span className="text-[var(--chrome-fg-dim)]">·</span>}
-            {info && <span>{modelBadge}</span>}
-            {info?.fast_download?.xet_enabled && (
-              <>
-                <span className="text-[var(--chrome-fg-dim)]">·</span>
-                <span
-                  className="text-[var(--chrome-accent)]"
-                  title={
-                    t('models.fast_download_title', {
-                      version: info.fast_download.xet_version || 'Xet',
-                    }) ||
-                    `Fast downloads via Xet ${info.fast_download.xet_version || ''} — parallel chunked transfer`
-                  }
-                >
-                  ⚡ {t('models.fast_download_badge') || 'fast download'}
-                </span>
-              </>
-            )}
-          </div>
-          <div className="inline-flex items-center gap-[var(--space-2)]">
-            {/* Compact HF token inline */}
-            {!hfTokenSet && !hfExpanded && (
-              <button
-                className="inline-flex cursor-pointer items-center gap-1 rounded-[var(--chrome-radius-pill)] [border:1px_solid_var(--chrome-border)] bg-transparent px-[var(--space-2)] py-[2px] text-[var(--chrome-fg-muted)] hover:bg-[var(--chrome-hover-bg)] hover:text-[var(--chrome-fg)]"
-                onClick={() => setHfExpanded(true)}
-                title={t('models.hf_set_title')}
-              >
-                <KeyRound size={11} /> {t('models.hf_token_btn')}
-              </button>
-            )}
-            {!hfTokenSet && hfExpanded && (
-              <div className="inline-flex items-center gap-[var(--space-2)]">
-                <input
-                  type="password"
-                  className="min-w-0 rounded-[var(--chrome-radius-pill)] [border:1px_solid_var(--chrome-border)] bg-[var(--chrome-hover-bg)] px-[var(--space-2)] py-[2px] font-[family-name:var(--chrome-font-mono)] text-[length:var(--text-xs)] text-[var(--chrome-fg)] placeholder:text-[var(--chrome-fg-dim)] focus-visible:border-[var(--chrome-accent)] focus-visible:shadow-[var(--focus-ring)] focus-visible:outline-none"
-                  placeholder="hf_xxxxxxxxxxxx"
-                  value={hfToken}
-                  onChange={(e) => setHfToken(e.target.value)}
-                  onKeyDown={(e) => {
-                    if (e.key === 'Enter') saveHfToken();
-                    if (e.key === 'Escape') setHfExpanded(false);
-                  }}
-                  autoFocus
-                />
-                <Button
-                  size="sm"
-                  variant="subtle"
-                  onClick={saveHfToken}
-                  disabled={hfSaving || !hfToken.trim()}
-                  loading={hfSaving}
-                >
-                  {t('common.save')}
-                </Button>
-                <a
-                  href="#"
-                  className="text-[var(--chrome-accent)] no-underline hover:underline"
-                  onClick={(e) => {
-                    e.preventDefault();
-                    openExternal('https://huggingface.co/settings/tokens');
-                  }}
-                  title="Open huggingface.co/settings/tokens"
-                >
-                  {t('models.get_token')}→
-                </a>
-              </div>
-            )}
-            {hfTokenSet && (
-              <span className="inline-flex items-center gap-1 text-[var(--chrome-severity-ok)]">
-                <KeyRound size={10} /> ✓
-              </span>
-            )}
-            <Button
-              variant="subtle"
-              size="sm"
-              onClick={reload}
-              loading={loading}
-              leading={<RefreshCw size={11} />}
-            >
-              {t('common.refresh')}
-            </Button>
-          </div>
-        </div>
-
-        <RecoBanner
-          reco={reco}
-          t={t}
-          installMutation={installMutation}
-          installingReco={installingReco}
-          setInstallingReco={setInstallingReco}
-          onInstallRecommended={onInstallRecommended}
-          onInstall={onInstall}
-          getRowRuntime={getRowRuntime}
-          diskFreeGb={data.disk_free_gb}
-        />
-
-        <div
-          className={`${catalogueLayout ? 'my-[24px]' : 'my-[var(--space-2)]'} flex items-center gap-[var(--space-2)] max-[580px]:flex-col max-[580px]:items-stretch`}
-        >
-          <SettingsInput
+    <section className="flex min-h-0 flex-col font-sans" data-testid="model-list-panel">
+      <div className="mb-[18px] flex flex-wrap items-center gap-[10px] px-[2px]">
+        <div className="relative w-full min-w-[160px] max-w-[420px]">
+          <Search
+            size={13}
+            className="pointer-events-none absolute left-[10px] top-1/2 -translate-y-1/2 text-muted-foreground"
+            aria-hidden="true"
+          />
+          <Input
             type="search"
-            className="max-w-none flex-1 text-[length:var(--text-xs)] min-w-[120px]"
+            className="h-8 pl-[30px] text-sm"
             placeholder={t('models.search_placeholder')}
             value={query}
             onChange={(e) => setQuery(e.target.value)}
             aria-label={t('models.search_label')}
           />
         </div>
-
-        <div className="min-h-0 flex-1" data-testid="model-list-area">
-          {sections.map((group) => (
-            <ModelSection
-              key={group.key}
-              sectionKey={group.key}
-              title={MODEL_SECTION_LABEL[group.key] || group.key}
-              group={group}
-              columns={columns}
-              getRowRuntime={getRowRuntime}
-              t={t}
-            />
-          ))}
-        </div>
-        {/* Global empty state — every section filtered out. Same actionable
-          "Clear filters" affordance the table-level empty state used to carry. */}
-        {sections.length === 0 && allModels.length > 0 && (
-          <div className="models-table__empty">
-            <span>{t('models.no_matches')}</span>
-            <Button
-              size="sm"
-              variant="subtle"
-              className="ml-[8px]"
-              onClick={() => setQuery('')}
-              data-testid="models-clear-filters"
-            >
-              {t('models.clear_filters')}
-            </Button>
-          </div>
+        {info?.fast_download?.xet_enabled && (
+          <span
+            className="inline-flex items-center gap-1 font-mono text-[11px] text-accent"
+            title={
+              t('models.fast_download_title', {
+                version: info.fast_download.xet_version || 'Xet',
+              }) ||
+              `Fast downloads via Xet ${info.fast_download.xet_version || ''} — parallel chunked transfer`
+            }
+          >
+            <Zap size={11} aria-hidden="true" />{' '}
+            {t('models.fast_download_badge') || 'fast download'}
+          </span>
         )}
-      </section>
-      {/* Downloaded previews belong next to downloaded models: same question
-        ("what has this install fetched?"), same tab. */}
-      <VoicePreviewsPanel />
-    </>
+        <span className="flex-1" />
+        <Button
+          variant="subtle"
+          size="sm"
+          onClick={reload}
+          loading={loading}
+          leading={<RefreshCw size={11} />}
+        >
+          {t('common.refresh')}
+        </Button>
+      </div>
+
+      <RecoBanner
+        reco={familyReco}
+        t={t}
+        installMutation={installMutation}
+        installingReco={installingReco}
+        setInstallingReco={setInstallingReco}
+        onInstallRecommended={onInstallRecommended}
+        onInstall={onInstall}
+        getRowRuntime={getRowRuntime}
+        diskFreeGb={data.disk_free_gb}
+      />
+
+      <div className="min-h-0 flex-1" data-testid="model-list-area">
+        {sections.map((group) => (
+          <ModelSection
+            key={group.key}
+            sectionKey={group.key}
+            title={MODEL_SECTION_LABEL[group.key] || group.key}
+            group={group}
+            columns={columns}
+            getRowRuntime={getRowRuntime}
+            t={t}
+          />
+        ))}
+      </div>
+      {/* Global empty state — every section filtered out. Same actionable
+          "Clear filters" affordance the table-level empty state carries. */}
+      {sections.length === 0 && allModels.length > 0 && (
+        <div className="flex items-center gap-2 px-[2px] py-[18px] text-sm text-muted-foreground">
+          <span>{t('models.no_matches')}</span>
+          <Button
+            size="sm"
+            variant="subtle"
+            onClick={() => setQuery('')}
+            data-testid="models-clear-filters"
+          >
+            {t('models.clear_filters')}
+          </Button>
+        </div>
+      )}
+    </section>
   );
 }

frontend/src/components/settings/SettingsSidebar.test.jsx

@@ -77,7 +77,7 @@ describe('SettingsSidebar — keyboard navigation', () => {
   it('is a single tab stop: only the active category is tabbable', () => {
     render(<SettingsSidebar active="appearance" onSelect={() => {}} />);
     expect(screen.getByTestId('settings-nav-appearance')).toHaveAttribute('tabindex', '0');
-    expect(screen.getByTestId('settings-nav-engines')).toHaveAttribute('tabindex', '-1');
+    expect(screen.getByTestId('settings-nav-dictation')).toHaveAttribute('tabindex', '-1');
   });
 
   it('ArrowDown/ArrowUp move selection, crossing group boundaries', () => {
@@ -86,12 +86,12 @@ describe('SettingsSidebar — keyboard navigation', () => {
     // the first item of Voice & Engines — arrowing must not stop at the seam.
     render(<SettingsSidebar active="appearance" onSelect={onSelect} />);
     fireEvent.keyDown(nav(), { key: 'ArrowDown' });
-    expect(onSelect).toHaveBeenCalledWith('engines');
+    expect(onSelect).toHaveBeenCalledWith('dictation');
 
     onSelect.mockClear();
-    render(<SettingsSidebar active="models" onSelect={onSelect} />);
+    render(<SettingsSidebar active="pronunciation" onSelect={onSelect} />);
     fireEvent.keyDown(screen.getAllByTestId('settings-nav-scroll')[1], { key: 'ArrowUp' });
-    expect(onSelect).toHaveBeenCalledWith('engines');
+    expect(onSelect).toHaveBeenCalledWith('dictation');
   });
 
   it('Home/End jump to the first and last visible category', () => {

frontend/src/components/settings/StorageUsagePanel.jsx

@@ -132,7 +132,7 @@ function SmallButton({ children, onClick, title, testId, disabled }) {
 
 export default function StorageUsagePanel() {
   const { t } = useTranslation();
-  const openSettingsTab = useAppStore((s) => s.openSettingsTab);
+  const openCatalogue = useAppStore((s) => s.openCatalogue);
   const [data, setData] = useState(null);
   const [loading, setLoading] = useState(true);
   const [refreshing, setRefreshing] = useState(false);
@@ -379,7 +379,7 @@ export default function StorageUsagePanel() {
                     </span>
                     {cat.id === 'hf_cache' && (
                       <SmallButton
-                        onClick={() => openSettingsTab('models')}
+                        onClick={() => openCatalogue()}
                         title={t('settings.storage_manage_models_hint', {
                           defaultValue: 'Reclaim space by removing models in the Model Store',
                         })}

frontend/src/components/settings/models/sections.js

@@ -68,3 +68,31 @@ export function groupModels(models, query) {
     };
   });
 }
+
+/** Which catalog sections belong to each engine family (one family's slice of the catalog). */
+export const FAMILY_SECTIONS = {
+  tts: ['tts'],
+  asr: ['asr', 'dictation', 'diarisation'],
+  llm: [],
+};
+
+/**
+ * Pure: narrow the device-wide recommendation preset to one family so the
+ * banner under a family's engines only talks about that family's weights.
+ * Totals are recomputed from the subset; null when nothing applies.
+ */
+export function scopeReco(reco, family) {
+  if (!reco || !family) return reco;
+  const keep = FAMILY_SECTIONS[family] || [];
+  const models = (reco.models || []).filter((m) => keep.includes(modelSectionKey(m)));
+  if (models.length === 0) return null;
+  const missing = models.filter((m) => !m.installed);
+  const gb = (xs) => Math.round(xs.reduce((sum, m) => sum + (m.size_gb || 0), 0) * 10) / 10;
+  return {
+    ...reco,
+    models,
+    all_installed: missing.length === 0,
+    total_gb: gb(models),
+    download_gb_remaining: gb(missing),
+  };
+}

frontend/src/components/settings/settingsCategories.jsx

@@ -90,49 +90,6 @@ export const GROUPS = [
     labelKey: 'settings.group_voice',
     defaultLabel: 'Voice & Engines',
     items: [
-      {
-        id: 'engines',
-        labelKey: 'settings.engines',
-        defaultLabel: 'Engines',
-        icon: Plug,
-        // Engine selection lives in the Model Catalogue workspace now; the
-        // category stays so the old name still finds it (it renders a pointer).
-        keywords: [
-          'engine',
-          'catalogue',
-          'catalog',
-          'model catalogue',
-          'default engine',
-          'tts engine',
-          'indextts',
-          'cosyvoice',
-          'compatibility',
-          'gpu',
-          'asr',
-          'transcription',
-          'whisper',
-          'openai-compatible',
-          'remote asr',
-        ],
-      },
-      {
-        id: 'models',
-        labelKey: 'settings.models',
-        defaultLabel: 'Models',
-        icon: Cpu,
-        restart: true,
-        keywords: [
-          'model',
-          'download',
-          'cache directory',
-          'models directory',
-          'hugging face mirror',
-          'hf_endpoint',
-          'catalogue',
-          'catalog',
-          'model store',
-        ],
-      },
       {
         id: 'dictation',
         labelKey: 'settings.dictation',
@@ -231,6 +188,8 @@ export const GROUPS = [
       },
       {
         id: 'storage',
+        // Hosts a restart-bound panel (models directory / HF mirror).
+        restart: true,
         labelKey: 'settings.storage',
         defaultLabel: 'Storage',
         icon: HardDrive,
@@ -251,6 +210,13 @@ export const GROUPS = [
           'free space',
           'disk space',
           'model cache size',
+          'models directory',
+          'cache directory',
+          'model store',
+          'catalogue',
+          'catalog',
+          'download',
+          'voice previews',
           'engine venvs',
           'temp files',
           'clear logs',
@@ -279,6 +245,8 @@ export const GROUPS = [
       },
       {
         id: 'network',
+        // Hosts a restart-bound panel (models directory / HF mirror).
+        restart: true,
         labelKey: 'settings.network',
         defaultLabel: 'Network',
         icon: Wifi,
@@ -297,6 +265,9 @@ export const GROUPS = [
         // settingsCategories.test.jsx like Models / Performance / Sharing.
         restart: true,
         keywords: [
+          'hugging face mirror',
+          'hf_endpoint',
+          'mirror',
           'ffmpeg',
           'ffprobe',
           'ffmpeg path',

frontend/src/components/settings/settingsCategories.test.jsx

@@ -62,8 +62,8 @@ describe('restart flag ↔ RestartBadge lockstep', () => {
   // restart: true, or the sidebar ↻ glyph / header badge contract breaks
   // (that drift is exactly how Network shipped without its glyph).
   const PANEL_CATEGORY = {
-    'StoragePanel.jsx': 'models',
-    'HFMirrorPanel.jsx': 'models',
+    'StoragePanel.jsx': 'storage',
+    'HFMirrorPanel.jsx': 'network',
     'RemoteBackendPanel.jsx': 'sharing',
     'AudioToolsPanel.jsx': 'audio-tools',
     'PerformancePanel.jsx': 'performance',

frontend/src/i18n/locales/ar.json

@@ -33,13 +33,16 @@
   },
   "catalogue": {
     "title": "كتالوج النماذج",
-    "tab_engines": "المحركات",
-    "tab_models": "النماذج",
-    "moved_title": "أصبح ضمن كتالوج النماذج",
-    "moved_engines_body": "انتقل اختيار المحرك الافتراضي للنطق والتفريغ النصي ونماذج اللغة الكبيرة إلى مساحة عمل كتالوج النماذج.",
-    "moved_models_body": "انتقل تصفح النماذج المنزَّلة وتثبيتها وإزالتها إلى مساحة عمل كتالوج النماذج.",
-    "open_engines": "فتح المحركات",
-    "open_models": "فتح النماذج"
+    "summary_title": "إعدادك",
+    "change": "تغيير",
+    "status_setup": "يحتاج إلى إعداد",
+    "status_cpu": "احتياطي CPU",
+    "status_off": "متوقف",
+    "weights": "الأوزان المنزَّلة",
+    "storage_link": "إعدادات التخزين",
+    "reco_line": "تم تثبيت {{installed}} من {{total}} نماذج موصى بها لـ {{device}} · يتبقى {{remaining}} GB",
+    "install_rest": "تثبيت الباقي",
+    "device_remote": "بعيد"
   },
   "settings": {
     "remote_backend_test": "اختبار الاتصال",
@@ -779,7 +782,7 @@
     "configureProviders": "إعداد مقدمي الخدمة",
     "loading": "جارٍ تحميل المحركات…",
     "refresh": "تحديث",
-    "matrixTitle": "مصفوفة توافق المحرك",
+    "matrixTitle": "المحركات",
     "loadFailed": "فشل تحميل المحركات: {{message}}",
     "couldNotLoad": "لا يمكن تحميل المحركات: {{message}}",
     "retry": "أعد المحاولة",

frontend/src/i18n/locales/de.json

@@ -33,13 +33,16 @@
   },
   "catalogue": {
     "title": "Modellkatalog",
-    "tab_engines": "Engines",
-    "tab_models": "Modelle",
-    "moved_title": "Jetzt im Modellkatalog",
-    "moved_engines_body": "Die Auswahl der Standard-Engine für Sprache, Transkription und LLM befindet sich jetzt im Arbeitsbereich „Modellkatalog“.",
-    "moved_models_body": "Heruntergeladene Modelle durchsehen, installieren und entfernen befindet sich jetzt im Arbeitsbereich „Modellkatalog“.",
-    "open_engines": "Engines öffnen",
-    "open_models": "Modelle öffnen"
+    "summary_title": "Deine Einrichtung",
+    "change": "Ändern",
+    "status_setup": "Einrichtung nötig",
+    "status_cpu": "CPU-Fallback",
+    "status_off": "Aus",
+    "weights": "Heruntergeladene Gewichte",
+    "storage_link": "Speichereinstellungen",
+    "reco_line": "{{installed}} von {{total}} empfohlenen Modellen für {{device}} installiert · noch {{remaining}} GB",
+    "install_rest": "Rest installieren",
+    "device_remote": "Remote"
   },
   "settings": {
     "remote_backend_test": "Verbindung testen",
@@ -779,7 +782,7 @@
     "configureProviders": "Anbieter konfigurieren",
     "loading": "Motoren werden geladen…",
     "refresh": "Aktualisieren",
-    "matrixTitle": "Engine-Kompatibilitätsmatrix",
+    "matrixTitle": "Engines",
     "loadFailed": "Fehler beim Laden der Engines: {{message}}",
     "couldNotLoad": "Engines konnten nicht geladen werden: {{message}}",
     "retry": "Versuchen Sie es noch einmal",

frontend/src/i18n/locales/en.json

@@ -171,13 +171,16 @@
   },
   "catalogue": {
     "title": "Model Catalogue",
-    "tab_engines": "Engines",
-    "tab_models": "Models",
-    "moved_title": "Now in the Model Catalogue",
-    "moved_engines_body": "Choosing the default speech, transcription and LLM engine moved to the Model Catalogue workspace.",
-    "moved_models_body": "Browsing, installing and removing model downloads moved to the Model Catalogue workspace.",
-    "open_engines": "Open Engines",
-    "open_models": "Open Models"
+    "summary_title": "Your setup",
+    "change": "Change",
+    "status_setup": "Needs setup",
+    "status_cpu": "CPU fallback",
+    "status_off": "Off",
+    "weights": "Downloaded weights",
+    "storage_link": "Storage settings",
+    "reco_line": "{{installed}} of {{total}} recommended models for {{device}} installed · {{remaining}} GB to go",
+    "install_rest": "Install the rest",
+    "device_remote": "Remote"
   },
   "audiobook": {
     "cast": "Cast",
@@ -2118,7 +2121,7 @@
     "configureProviders": "Configure providers",
     "loading": "Loading engines…",
     "refresh": "Refresh",
-    "matrixTitle": "Engine Compatibility Matrix",
+    "matrixTitle": "Engines",
     "familyMatrixTitle": "{{family}} Engines",
     "colEngine": "Engine",
     "colGpuCompat": "GPU compat",

frontend/src/i18n/locales/es.json

@@ -33,13 +33,16 @@
   },
   "catalogue": {
     "title": "Catálogo de modelos",
-    "tab_engines": "Motores",
-    "tab_models": "Modelos",
-    "moved_title": "Ahora en el catálogo de modelos",
-    "moved_engines_body": "Elegir el motor predeterminado de voz, transcripción y LLM se movió al espacio Catálogo de modelos.",
-    "moved_models_body": "Explorar, instalar y eliminar descargas de modelos se movió al espacio Catálogo de modelos.",
-    "open_engines": "Abrir motores",
-    "open_models": "Abrir modelos"
+    "summary_title": "Tu configuración",
+    "change": "Cambiar",
+    "status_setup": "Requiere configuración",
+    "status_cpu": "Alternativa CPU",
+    "status_off": "Desactivado",
+    "weights": "Pesos descargados",
+    "storage_link": "Ajustes de almacenamiento",
+    "reco_line": "{{installed}} de {{total}} modelos recomendados para {{device}} instalados · faltan {{remaining}} GB",
+    "install_rest": "Instalar el resto",
+    "device_remote": "Remoto"
   },
   "settings": {
     "remote_backend_test": "Probar conexión",
@@ -779,7 +782,7 @@
     "configureProviders": "Configurar proveedores",
     "loading": "Cargando motores…",
     "refresh": "Actualizar",
-    "matrixTitle": "Matriz de compatibilidad de motores",
+    "matrixTitle": "Motores",
     "loadFailed": "No se pudieron cargar los motores: {{message}}",
     "couldNotLoad": "No se pudieron cargar los motores: {{message}}",
     "retry": "Reintentar",

frontend/src/i18n/locales/fr.json

@@ -33,13 +33,16 @@
   },
   "catalogue": {
     "title": "Catalogue de modèles",
-    "tab_engines": "Moteurs",
-    "tab_models": "Modèles",
-    "moved_title": "Désormais dans le catalogue de modèles",
-    "moved_engines_body": "Le choix du moteur par défaut pour la voix, la transcription et les LLM se trouve maintenant dans l’espace Catalogue de modèles.",
-    "moved_models_body": "Parcourir, installer et supprimer les modèles téléchargés se trouve maintenant dans l’espace Catalogue de modèles.",
-    "open_engines": "Ouvrir les moteurs",
-    "open_models": "Ouvrir les modèles"
+    "summary_title": "Votre configuration",
+    "change": "Modifier",
+    "status_setup": "Configuration requise",
+    "status_cpu": "Repli CPU",
+    "status_off": "Désactivé",
+    "weights": "Poids téléchargés",
+    "storage_link": "Paramètres de stockage",
+    "reco_line": "{{installed}} des {{total}} modèles recommandés pour {{device}} installés · {{remaining}} Go restants",
+    "install_rest": "Installer le reste",
+    "device_remote": "Distant"
   },
   "settings": {
     "remote_backend_test": "Tester la connexion",
@@ -779,7 +782,7 @@
     "configureProviders": "Configurer les fournisseurs",
     "loading": "Chargement des moteurs…",
     "refresh": "Actualiser",
-    "matrixTitle": "Matrice de compatibilité des moteurs",
+    "matrixTitle": "Moteurs",
     "loadFailed": "Échec du chargement des moteurs : {{message}}",
     "couldNotLoad": "Impossible de charger les moteurs : {{message}}",
     "retry": "Réessayer",

frontend/src/i18n/locales/hi.json

@@ -33,13 +33,16 @@
   },
   "catalogue": {
     "title": "मॉडल कैटलॉग",
-    "tab_engines": "इंजन",
-    "tab_models": "मॉडल",
-    "moved_title": "अब मॉडल कैटलॉग में",
-    "moved_engines_body": "वाक्, प्रतिलेखन और LLM का डिफ़ॉल्ट इंजन चुनना अब मॉडल कैटलॉग वर्कस्पेस में है।",
-    "moved_models_body": "डाउनलोड किए गए मॉडल देखना, इंस्टॉल करना और हटाना अब मॉडल कैटलॉग वर्कस्पेस में है।",
-    "open_engines": "इंजन खोलें",
-    "open_models": "मॉडल खोलें"
+    "summary_title": "आपका सेटअप",
+    "change": "बदलें",
+    "status_setup": "सेटअप आवश्यक",
+    "status_cpu": "CPU फ़ॉलबैक",
+    "status_off": "बंद",
+    "weights": "डाउनलोड किए गए वेट्स",
+    "storage_link": "स्टोरेज सेटिंग्स",
+    "reco_line": "{{device}} के लिए {{total}} में से {{installed}} अनुशंसित मॉडल इंस्टॉल · {{remaining}} GB शेष",
+    "install_rest": "बाकी इंस्टॉल करें",
+    "device_remote": "रिमोट"
   },
   "settings": {
     "remote_backend_test": "कनेक्शन जाँचें",
@@ -779,7 +782,7 @@
     "configureProviders": "प्रदाता कॉन्फ़िगर करें",
     "loading": "इंजन लोड हो रहे हैं...",
     "refresh": "ताज़ा करें",
-    "matrixTitle": "इंजन अनुकूलता मैट्रिक्स",
+    "matrixTitle": "इंजन",
     "loadFailed": "इंजन लोड करने में विफल: {{message}}",
     "couldNotLoad": "इंजन लोड नहीं हो सका: {{message}}",
     "retry": "पुनः प्रयास करें",

frontend/src/i18n/locales/id.json

@@ -33,13 +33,16 @@
   },
   "catalogue": {
     "title": "Katalog model",
-    "tab_engines": "Engine",
-    "tab_models": "Model",
-    "moved_title": "Kini di Katalog model",
-    "moved_engines_body": "Pemilihan engine default untuk suara, transkripsi, dan LLM kini ada di ruang kerja Katalog model.",
-    "moved_models_body": "Menjelajahi, memasang, dan menghapus model yang diunduh kini ada di ruang kerja Katalog model.",
-    "open_engines": "Buka engine",
-    "open_models": "Buka model"
+    "summary_title": "Pengaturan Anda",
+    "change": "Ubah",
+    "status_setup": "Perlu penyiapan",
+    "status_cpu": "Fallback CPU",
+    "status_off": "Nonaktif",
+    "weights": "Bobot yang diunduh",
+    "storage_link": "Pengaturan penyimpanan",
+    "reco_line": "{{installed}} dari {{total}} model yang direkomendasikan untuk {{device}} terpasang · sisa {{remaining}} GB",
+    "install_rest": "Pasang sisanya",
+    "device_remote": "Jarak jauh"
   },
   "settings": {
     "remote_backend_test": "Uji koneksi",
@@ -779,7 +782,7 @@
     "configureProviders": "Konfigurasikan penyedia",
     "loading": "Memuat mesin…",
     "refresh": "Segarkan",
-    "matrixTitle": "Matriks Kompatibilitas Mesin",
+    "matrixTitle": "Mesin",
     "loadFailed": "Gagal memuat mesin: {{message}}",
     "couldNotLoad": "Tidak dapat memuat mesin: {{message}}",
     "retry": "Coba lagi",

frontend/src/i18n/locales/it.json

@@ -33,13 +33,16 @@
   },
   "catalogue": {
     "title": "Catalogo modelli",
-    "tab_engines": "Motori",
-    "tab_models": "Modelli",
-    "moved_title": "Ora nel catalogo modelli",
-    "moved_engines_body": "La scelta del motore predefinito per voce, trascrizione e LLM si trova ora nello spazio Catalogo modelli.",
-    "moved_models_body": "Sfogliare, installare e rimuovere i modelli scaricati si trova ora nello spazio Catalogo modelli.",
-    "open_engines": "Apri motori",
-    "open_models": "Apri modelli"
+    "summary_title": "La tua configurazione",
+    "change": "Cambia",
+    "status_setup": "Da configurare",
+    "status_cpu": "Fallback CPU",
+    "status_off": "Disattivato",
+    "weights": "Pesi scaricati",
+    "storage_link": "Impostazioni archiviazione",
+    "reco_line": "{{installed}} di {{total}} modelli consigliati per {{device}} installati · mancano {{remaining}} GB",
+    "install_rest": "Installa il resto",
+    "device_remote": "Remoto"
   },
   "settings": {
     "remote_backend_test": "Verifica connessione",
@@ -779,7 +782,7 @@
     "configureProviders": "Configura i provider",
     "loading": "Caricamento motori…",
     "refresh": "Aggiorna",
-    "matrixTitle": "Matrice di compatibilità del motore",
+    "matrixTitle": "Motori",
     "loadFailed": "Impossibile caricare i motori: {{message}}",
     "couldNotLoad": "Impossibile caricare i motori: {{message}}",
     "retry": "Riprova",

frontend/src/i18n/locales/ja.json

@@ -33,13 +33,16 @@
   },
   "catalogue": {
     "title": "モデルカタログ",
-    "tab_engines": "エンジン",
-    "tab_models": "モデル",
-    "moved_title": "モデルカタログに移動しました",
-    "moved_engines_body": "音声・文字起こし・LLM の既定エンジンの選択は「モデルカタログ」ワークスペースに移動しました。",
-    "moved_models_body": "モデルの閲覧・インストール・削除は「モデルカタログ」ワークスペースに移動しました。",
-    "open_engines": "エンジンを開く",
-    "open_models": "モデルを開く"
+    "summary_title": "現在の設定",
+    "change": "変更",
+    "status_setup": "設定が必要",
+    "status_cpu": "CPUフォールバック",
+    "status_off": "オフ",
+    "weights": "ダウンロード済みの重み",
+    "storage_link": "ストレージ設定",
+    "reco_line": "{{device}} 向けの推奨モデル {{total}} 件中 {{installed}} 件をインストール済み · 残り {{remaining}} GB",
+    "install_rest": "残りをインストール",
+    "device_remote": "リモート"
   },
   "settings": {
     "remote_backend_test": "接続をテスト",
@@ -779,7 +782,7 @@
     "configureProviders": "プロバイダーを設定",
     "loading": "エンジンをロード中…",
     "refresh": "リフレッシュ",
-    "matrixTitle": "エンジン互換性マトリックス",
+    "matrixTitle": "エンジン",
     "loadFailed": "エンジンのロードに失敗しました: {{message}}",
     "couldNotLoad": "エンジンをロードできませんでした: {{message}}",
     "retry": "再試行",

frontend/src/i18n/locales/ko.json

@@ -33,13 +33,16 @@
   },
   "catalogue": {
     "title": "모델 카탈로그",
-    "tab_engines": "엔진",
-    "tab_models": "모델",
-    "moved_title": "이제 모델 카탈로그에 있습니다",
-    "moved_engines_body": "음성, 전사, LLM 기본 엔진 선택이 모델 카탈로그 작업 공간으로 이동했습니다.",
-    "moved_models_body": "내려받은 모델을 살펴보고 설치·삭제하는 기능이 모델 카탈로그 작업 공간으로 이동했습니다.",
-    "open_engines": "엔진 열기",
-    "open_models": "모델 열기"
+    "summary_title": "내 설정",
+    "change": "변경",
+    "status_setup": "설정 필요",
+    "status_cpu": "CPU 대체",
+    "status_off": "꺼짐",
+    "weights": "다운로드한 가중치",
+    "storage_link": "저장소 설정",
+    "reco_line": "{{device}} 권장 모델 {{total}}개 중 {{installed}}개 설치됨 · {{remaining}} GB 남음",
+    "install_rest": "나머지 설치",
+    "device_remote": "원격"
   },
   "settings": {
     "remote_backend_test": "연결 테스트",
@@ -1023,7 +1026,7 @@
     "configureProviders": "제공업체 설정",
     "loading": "엔진 로드 중…",
     "refresh": "새로고침",
-    "matrixTitle": "엔진 호환성 매트릭스",
+    "matrixTitle": "엔진",
     "loadFailed": "엔진을 로드하지 못했습니다: {{message}}",
     "couldNotLoad": "엔진을 로드할 수 없습니다: {{message}}",
     "retry": "재시도",

frontend/src/i18n/locales/nl.json

@@ -33,13 +33,16 @@
   },
   "catalogue": {
     "title": "Modelcatalogus",
-    "tab_engines": "Engines",
-    "tab_models": "Modellen",
-    "moved_title": "Nu in de modelcatalogus",
-    "moved_engines_body": "Het kiezen van de standaard-engine voor spraak, transcriptie en LLM staat nu in de werkruimte Modelcatalogus.",
-    "moved_models_body": "Gedownloade modellen bekijken, installeren en verwijderen staat nu in de werkruimte Modelcatalogus.",
-    "open_engines": "Engines openen",
-    "open_models": "Modellen openen"
+    "summary_title": "Jouw configuratie",
+    "change": "Wijzigen",
+    "status_setup": "Instellen vereist",
+    "status_cpu": "CPU-fallback",
+    "status_off": "Uit",
+    "weights": "Gedownloade gewichten",
+    "storage_link": "Opslaginstellingen",
+    "reco_line": "{{installed}} van {{total}} aanbevolen modellen voor {{device}} geïnstalleerd · nog {{remaining}} GB",
+    "install_rest": "Rest installeren",
+    "device_remote": "Extern"
   },
   "settings": {
     "remote_backend_test": "Verbinding testen",
@@ -779,7 +782,7 @@
     "configureProviders": "Providers configureren",
     "loading": "Motoren laden…",
     "refresh": "Vernieuwen",
-    "matrixTitle": "Motorcompatibiliteitsmatrix",
+    "matrixTitle": "Motoren",
     "loadFailed": "Kan motoren niet laden: {{message}}",
     "couldNotLoad": "Kan motoren niet laden: {{message}}",
     "retry": "Opnieuw proberen",

frontend/src/i18n/locales/pl.json

@@ -33,13 +33,16 @@
   },
   "catalogue": {
     "title": "Katalog modeli",
-    "tab_engines": "Silniki",
-    "tab_models": "Modele",
-    "moved_title": "Teraz w katalogu modeli",
-    "moved_engines_body": "Wybór domyślnego silnika mowy, transkrypcji i LLM znajduje się teraz w obszarze Katalog modeli.",
-    "moved_models_body": "Przeglądanie, instalowanie i usuwanie pobranych modeli znajduje się teraz w obszarze Katalog modeli.",
-    "open_engines": "Otwórz silniki",
-    "open_models": "Otwórz modele"
+    "summary_title": "Twoja konfiguracja",
+    "change": "Zmień",
+    "status_setup": "Wymaga konfiguracji",
+    "status_cpu": "Tryb awaryjny CPU",
+    "status_off": "Wyłączone",
+    "weights": "Pobrane wagi",
+    "storage_link": "Ustawienia pamięci",
+    "reco_line": "Zainstalowano {{installed}} z {{total}} zalecanych modeli dla {{device}} · pozostało {{remaining}} GB",
+    "install_rest": "Zainstaluj resztę",
+    "device_remote": "Zdalny"
   },
   "settings": {
     "remote_backend_test": "Testuj połączenie",
@@ -779,7 +782,7 @@
     "configureProviders": "Skonfiguruj dostawców",
     "loading": "Ładowanie silników…",
     "refresh": "Odśwież",
-    "matrixTitle": "Matryca kompatybilności silników",
+    "matrixTitle": "Silniki",
     "loadFailed": "Nie udało się załadować silników: {{message}}",
     "couldNotLoad": "Nie można załadować silników: {{message}}",
     "retry": "Spróbuj ponownie",

frontend/src/i18n/locales/pt.json

@@ -33,13 +33,16 @@
   },
   "catalogue": {
     "title": "Catálogo de modelos",
-    "tab_engines": "Motores",
-    "tab_models": "Modelos",
-    "moved_title": "Agora no catálogo de modelos",
-    "moved_engines_body": "A escolha do motor padrão de voz, transcrição e LLM foi para o espaço Catálogo de modelos.",
-    "moved_models_body": "Explorar, instalar e remover modelos baixados foi para o espaço Catálogo de modelos.",
-    "open_engines": "Abrir motores",
-    "open_models": "Abrir modelos"
+    "summary_title": "Sua configuração",
+    "change": "Alterar",
+    "status_setup": "Requer configuração",
+    "status_cpu": "Alternativa CPU",
+    "status_off": "Desligado",
+    "weights": "Pesos baixados",
+    "storage_link": "Configurações de armazenamento",
+    "reco_line": "{{installed}} de {{total}} modelos recomendados para {{device}} instalados · faltam {{remaining}} GB",
+    "install_rest": "Instalar o restante",
+    "device_remote": "Remoto"
   },
   "settings": {
     "remote_backend_test": "Testar conexão",
@@ -779,7 +782,7 @@
     "configureProviders": "Configurar provedores",
     "loading": "Carregando motores…",
     "refresh": "Atualizar",
-    "matrixTitle": "Matriz de compatibilidade do motor",
+    "matrixTitle": "Motores",
     "loadFailed": "Falha ao carregar motores: {{message}}",
     "couldNotLoad": "Não foi possível carregar os motores: {{message}}",
     "retry": "Tentar novamente",

frontend/src/i18n/locales/ru.json

@@ -33,13 +33,16 @@
   },
   "catalogue": {
     "title": "Каталог моделей",
-    "tab_engines": "Движки",
-    "tab_models": "Модели",
-    "moved_title": "Теперь в каталоге моделей",
-    "moved_engines_body": "Выбор движка по умолчанию для синтеза речи, расшифровки и LLM перенесён в раздел «Каталог моделей».",
-    "moved_models_body": "Просмотр, установка и удаление загруженных моделей перенесены в раздел «Каталог моделей».",
-    "open_engines": "Открыть движки",
-    "open_models": "Открыть модели"
+    "summary_title": "Ваша конфигурация",
+    "change": "Изменить",
+    "status_setup": "Требуется настройка",
+    "status_cpu": "Резерв на CPU",
+    "status_off": "Выкл.",
+    "weights": "Загруженные веса",
+    "storage_link": "Настройки хранилища",
+    "reco_line": "Установлено {{installed}} из {{total}} рекомендуемых моделей для {{device}} · осталось {{remaining}} ГБ",
+    "install_rest": "Установить остальное",
+    "device_remote": "Удалённый"
   },
   "settings": {
     "remote_backend_test": "Проверить подключение",
@@ -779,7 +782,7 @@
     "configureProviders": "Настроить поставщиков",
     "loading": "Загрузка двигателей…",
     "refresh": "Обновить",
-    "matrixTitle": "Матрица совместимости двигателей",
+    "matrixTitle": "Движки",
     "loadFailed": "Не удалось загрузить двигатели: {{message}}.",
     "couldNotLoad": "Не удалось загрузить двигатели: {{message}}",
     "retry": "Повторить попытку",

frontend/src/i18n/locales/sv.json

@@ -33,13 +33,16 @@
   },
   "catalogue": {
     "title": "Modellkatalog",
-    "tab_engines": "Motorer",
-    "tab_models": "Modeller",
-    "moved_title": "Nu i modellkatalogen",
-    "moved_engines_body": "Att välja standardmotor för tal, transkribering och LLM finns nu i arbetsytan Modellkatalog.",
-    "moved_models_body": "Att bläddra bland, installera och ta bort nedladdade modeller finns nu i arbetsytan Modellkatalog.",
-    "open_engines": "Öppna motorer",
-    "open_models": "Öppna modeller"
+    "summary_title": "Din konfiguration",
+    "change": "Ändra",
+    "status_setup": "Behöver konfigureras",
+    "status_cpu": "CPU-reserv",
+    "status_off": "Av",
+    "weights": "Nedladdade vikter",
+    "storage_link": "Lagringsinställningar",
+    "reco_line": "{{installed}} av {{total}} rekommenderade modeller för {{device}} installerade · {{remaining}} GB kvar",
+    "install_rest": "Installera resten",
+    "device_remote": "Fjärr"
   },
   "settings": {
     "remote_backend_test": "Testa anslutningen",
@@ -779,7 +782,7 @@
     "configureProviders": "Konfigurera leverantörer",
     "loading": "Laddar motorer...",
     "refresh": "Uppdatera",
-    "matrixTitle": "Motorkompatibilitetsmatris",
+    "matrixTitle": "Motorer",
     "loadFailed": "Det gick inte att ladda motorer: {{message}}",
     "couldNotLoad": "Kunde inte ladda motorer: {{message}}",
     "retry": "Försök igen",

frontend/src/i18n/locales/th.json

@@ -33,13 +33,16 @@
   },
   "catalogue": {
     "title": "แค็ตตาล็อกโมเดล",
-    "tab_engines": "เอนจิน",
-    "tab_models": "โมเดล",
-    "moved_title": "ย้ายไปที่แค็ตตาล็อกโมเดลแล้ว",
-    "moved_engines_body": "การเลือกเอนจินเริ่มต้นสำหรับเสียงพูด การถอดเสียง และ LLM ย้ายไปอยู่ในพื้นที่ทำงานแค็ตตาล็อกโมเดลแล้ว",
-    "moved_models_body": "การเรียกดู ติดตั้ง และลบโมเดลที่ดาวน์โหลดไว้ ย้ายไปอยู่ในพื้นที่ทำงานแค็ตตาล็อกโมเดลแล้ว",
-    "open_engines": "เปิดเอนจิน",
-    "open_models": "เปิดโมเดล"
+    "summary_title": "การตั้งค่าของคุณ",
+    "change": "เปลี่ยน",
+    "status_setup": "ต้องตั้งค่า",
+    "status_cpu": "สำรองด้วย CPU",
+    "status_off": "ปิด",
+    "weights": "น้ำหนักโมเดลที่ดาวน์โหลด",
+    "storage_link": "การตั้งค่าพื้นที่เก็บข้อมูล",
+    "reco_line": "ติดตั้งแล้ว {{installed}} จาก {{total}} โมเดลที่แนะนำสำหรับ {{device}} · เหลืออีก {{remaining}} GB",
+    "install_rest": "ติดตั้งส่วนที่เหลือ",
+    "device_remote": "ระยะไกล"
   },
   "settings": {
     "remote_backend_test": "ทดสอบการเชื่อมต่อ",
@@ -779,7 +782,7 @@
     "configureProviders": "ตั้งค่าผู้ให้บริการ",
     "loading": "กำลังโหลดเครื่องยนต์...",
     "refresh": "รีเฟรช",
-    "matrixTitle": "เมทริกซ์ความเข้ากันได้ของเครื่องยนต์",
+    "matrixTitle": "เอนจิน",
     "loadFailed": "ไม่สามารถโหลดเอ็นจิ้น: {{message}}",
     "couldNotLoad": "ไม่สามารถโหลดเอ็นจิ้น: {{message}}",
     "retry": "ลองอีกครั้ง",

frontend/src/i18n/locales/tr.json

@@ -33,13 +33,16 @@
   },
   "catalogue": {
     "title": "Model kataloğu",
-    "tab_engines": "Motorlar",
-    "tab_models": "Modeller",
-    "moved_title": "Artık model kataloğunda",
-    "moved_engines_body": "Varsayılan ses, deşifre ve LLM motorunu seçme işlemi Model kataloğu çalışma alanına taşındı.",
-    "moved_models_body": "İndirilen modellere göz atma, yükleme ve kaldırma işlemleri Model kataloğu çalışma alanına taşındı.",
-    "open_engines": "Motorları aç",
-    "open_models": "Modelleri aç"
+    "summary_title": "Kurulumunuz",
+    "change": "Değiştir",
+    "status_setup": "Kurulum gerekli",
+    "status_cpu": "CPU yedeği",
+    "status_off": "Kapalı",
+    "weights": "İndirilen ağırlıklar",
+    "storage_link": "Depolama ayarları",
+    "reco_line": "{{device}} için önerilen {{total}} modelden {{installed}} tanesi yüklü · {{remaining}} GB kaldı",
+    "install_rest": "Kalanları yükle",
+    "device_remote": "Uzak"
   },
   "settings": {
     "remote_backend_test": "Bağlantıyı test et",
@@ -779,7 +782,7 @@
     "configureProviders": "Sağlayıcıları yapılandır",
     "loading": "Motorlar yükleniyor…",
     "refresh": "Yenile",
-    "matrixTitle": "Motor Uyumluluk Matrisi",
+    "matrixTitle": "Motorlar",
     "loadFailed": "Motorlar yüklenemedi: {{message}}",
     "couldNotLoad": "Motorlar yüklenemedi: {{message}}",
     "retry": "Yeniden dene",

frontend/src/i18n/locales/uk.json

@@ -33,13 +33,16 @@
   },
   "catalogue": {
     "title": "Каталог моделей",
-    "tab_engines": "Рушії",
-    "tab_models": "Моделі",
-    "moved_title": "Тепер у каталозі моделей",
-    "moved_engines_body": "Вибір типового рушія мовлення, розшифрування та LLM перенесено до розділу «Каталог моделей».",
-    "moved_models_body": "Перегляд, встановлення та видалення завантажених моделей перенесено до розділу «Каталог моделей».",
-    "open_engines": "Відкрити рушії",
-    "open_models": "Відкрити моделі"
+    "summary_title": "Ваша конфігурація",
+    "change": "Змінити",
+    "status_setup": "Потрібне налаштування",
+    "status_cpu": "Резерв на CPU",
+    "status_off": "Вимкнено",
+    "weights": "Завантажені ваги",
+    "storage_link": "Налаштування сховища",
+    "reco_line": "Встановлено {{installed}} з {{total}} рекомендованих моделей для {{device}} · залишилось {{remaining}} ГБ",
+    "install_rest": "Встановити решту",
+    "device_remote": "Віддалений"
   },
   "settings": {
     "remote_backend_test": "Перевірити з’єднання",
@@ -779,7 +782,7 @@
     "configureProviders": "Налаштувати постачальників",
     "loading": "Завантаження двигунів…",
     "refresh": "Оновити",
-    "matrixTitle": "Матриця сумісності двигуна",
+    "matrixTitle": "Движки",
     "loadFailed": "Не вдалося завантажити двигуни: {{message}}",
     "couldNotLoad": "Не вдалося завантажити двигуни: {{message}}",
     "retry": "Повторіть спробу",

frontend/src/i18n/locales/vi.json

@@ -33,13 +33,16 @@
   },
   "catalogue": {
     "title": "Danh mục mô hình",
-    "tab_engines": "Engine",
-    "tab_models": "Mô hình",
-    "moved_title": "Đã chuyển sang Danh mục mô hình",
-    "moved_engines_body": "Việc chọn engine mặc định cho giọng nói, chuyển lời nói thành văn bản và LLM đã chuyển sang không gian làm việc Danh mục mô hình.",
-    "moved_models_body": "Việc duyệt, cài đặt và xoá mô hình đã tải đã chuyển sang không gian làm việc Danh mục mô hình.",
-    "open_engines": "Mở engine",
-    "open_models": "Mở mô hình"
+    "summary_title": "Thiết lập của bạn",
+    "change": "Thay đổi",
+    "status_setup": "Cần thiết lập",
+    "status_cpu": "Dự phòng CPU",
+    "status_off": "Tắt",
+    "weights": "Trọng số đã tải",
+    "storage_link": "Cài đặt lưu trữ",
+    "reco_line": "Đã cài {{installed}}/{{total}} mô hình đề xuất cho {{device}} · còn {{remaining}} GB",
+    "install_rest": "Cài phần còn lại",
+    "device_remote": "Từ xa"
   },
   "settings": {
     "remote_backend_test": "Kiểm tra kết nối",
@@ -779,7 +782,7 @@
     "configureProviders": "Cấu hình nhà cung cấp",
     "loading": "Đang tải động cơ…",
     "refresh": "Làm mới",
-    "matrixTitle": "Ma trận tương thích động cơ",
+    "matrixTitle": "Công cụ",
     "loadFailed": "Không tải được động cơ: {{message}}",
     "couldNotLoad": "Không thể tải động cơ: {{message}}",
     "retry": "Thử lại",

frontend/src/i18n/locales/zh-CN.json

@@ -152,13 +152,16 @@
   },
   "catalogue": {
     "title": "模型库",
-    "tab_engines": "引擎",
-    "tab_models": "模型",
-    "moved_title": "已移至模型库",
-    "moved_engines_body": "选择默认的语音合成、语音转写和大模型引擎已移至“模型库”工作区。",
-    "moved_models_body": "浏览、安装和删除已下载模型已移至“模型库”工作区。",
-    "open_engines": "打开引擎",
-    "open_models": "打开模型"
+    "summary_title": "当前配置",
+    "change": "更改",
+    "status_setup": "需要设置",
+    "status_cpu": "CPU 回退",
+    "status_off": "关闭",
+    "weights": "已下载的模型权重",
+    "storage_link": "存储设置",
+    "reco_line": "{{device}} 的推荐模型已安装 {{installed}}/{{total}} · 还需 {{remaining}} GB",
+    "install_rest": "安装其余模型",
+    "device_remote": "远程"
   },
   "launchpad": {
     "greeting": "你好",
@@ -975,7 +978,7 @@
     "configureProviders": "配置提供商",
     "loading": "加载引擎中…",
     "refresh": "刷新",
-    "matrixTitle": "引擎兼容性矩阵",
+    "matrixTitle": "引擎",
     "loadFailed": "无法加载引擎：{{message}}",
     "couldNotLoad": "无法加载引擎：{{message}}",
     "retry": "重试",

frontend/src/i18n/locales/zh-TW.json

@@ -33,13 +33,16 @@
   },
   "catalogue": {
     "title": "模型庫",
-    "tab_engines": "引擎",
-    "tab_models": "模型",
-    "moved_title": "已移至模型庫",
-    "moved_engines_body": "選擇預設的語音合成、語音轉寫與大型語言模型引擎已移至「模型庫」工作區。",
-    "moved_models_body": "瀏覽、安裝與移除已下載模型已移至「模型庫」工作區。",
-    "open_engines": "開啟引擎",
-    "open_models": "開啟模型"
+    "summary_title": "目前設定",
+    "change": "變更",
+    "status_setup": "需要設定",
+    "status_cpu": "CPU 回退",
+    "status_off": "關閉",
+    "weights": "已下載的模型權重",
+    "storage_link": "儲存設定",
+    "reco_line": "{{device}} 的推薦模型已安裝 {{installed}}/{{total}} · 尚需 {{remaining}} GB",
+    "install_rest": "安裝其餘模型",
+    "device_remote": "遠端"
   },
   "settings": {
     "remote_backend_test": "測試連線",
@@ -779,7 +782,7 @@
     "configureProviders": "設定提供者",
     "loading": "正在載入引擎...",
     "refresh": "重新整理",
-    "matrixTitle": "引擎相容性矩陣",
+    "matrixTitle": "配音引擎",
     "loadFailed": "無法載入引擎：{{message}}",
     "couldNotLoad": "無法載入引擎：{{message}}",
     "retry": "重試",

frontend/src/pages/ModelCatalogue.jsx

@@ -1,48 +1,43 @@
 /**
- * ModelCatalogue — the workspace where engines and model weights are browsed
- * and the app's defaults are chosen.
+ * ModelCatalogue — one page, one axis.
  *
- * Engine picking and the model store used to be two Settings categories, which
- * buried the single most consequential decision in the app (which TTS/ASR/LLM
- * engine runs, and which weights are on disk) three clicks deep, split across
- * two panes that constantly cross-reference each other. This promotes both to a
- * first-class workspace with one pane switch between them; Settings keeps only
- * the genuinely settings-shaped remainder (models directory, HF mirror) and
- * points here for the rest.
+ * Reads top-down the way a user thinks about it:
+ *   1. SetupSummary — what speech, transcription, dictation and the language
+ *      model use right now, one line each, with a Change button.
+ *   2. The engine list for ONE family (TTS / ASR / LLM), the same tested
+ *      EngineCompatibilityMatrix the Engines pane hosted; Change on a summary
+ *      row switches this family and scrolls here.
+ *   3. That family's downloadable weights (the former Models pane, filtered):
+ *      TTS weights under TTS; offline ASR, streaming dictation and diarisation
+ *      under ASR. LLM engines bring their own weights, so nothing renders there.
+ *   4. One storage line pointing at Settings → Storage.
  *
- * Deliberately a COMPOSITION, not a rewrite: the panes mount the existing
- * `EnginesTab` (engine matrix + the OpenAI-compatible ASR config) and
- * `ModelStoreTab` unchanged, so their data contracts, tests and behaviour carry
- * over untouched — the shared engine cache, the same install/delete flows, and
- * the same env-var-wins semantics.
+ * The previous Engines | Models pane switch put the same decision on two axes
+ * (family tabs inside one pane, role sections inside the other), duplicated
+ * dictation in both, and parked storage stats, the HF token and the voice
+ * preview toggle on the model list. Those now live where they belong
+ * (Settings → Storage / Credentials), and there is no pane to pick.
  *
- * `pendingCatalogueTab` is the one-shot deep-link hand-off (mirrors Settings'
- * `pendingSettingsTab`): a caller sets the pane and navigates here, this page
- * consumes it once and clears it so a later plain visit reopens the last pane.
+ * `pendingCatalogueFamily` is the one-shot deep-link hand-off (mirrors
+ * Settings' `pendingSettingsTab`): a caller names a family and navigates
+ * here; this page consumes it once and clears it so a later plain visit
+ * reopens the last family.
  */
-import React, { useCallback, useEffect, useMemo, useState } from 'react';
-import { Boxes, CheckCircle, Cpu, HardDriveDownload, RefreshCw } from 'lucide-react';
+import React, { useCallback, useEffect, useRef, useState } from 'react';
+import { Boxes } from 'lucide-react';
 import { useTranslation } from 'react-i18next';
 import { useAppStore } from '../store';
-import { useModelStatus, useSystemInfo } from '../api/hooks';
-import { Badge, Tabs } from '../ui';
+import { useModels } from '../api/hooks';
+import SetupSummary from '../components/catalogue/SetupSummary';
 import EnginesTab from '../components/settings/EnginesTab';
 import ModelStoreTab from '../components/settings/ModelStoreTab';
+import { fmtBytes } from '../components/settings/models/format';
 
 /** Persisted across visits so the workspace reopens where you left it. */
-const PANE_KEY = 'omnivoice.catalogue.pane';
 const FAMILY_KEY = 'omnivoice.catalogue.engine-family';
-const PANES = ['engines', 'models'];
 const FAMILIES = ['tts', 'asr', 'llm'];
-
-function readStoredPane() {
-  try {
-    const stored = localStorage.getItem(PANE_KEY);
-    return PANES.includes(stored) ? stored : 'engines';
-  } catch {
-    return 'engines';
-  }
-}
+/** Families whose engines load catalogue weights (LLM engines bring their own). */
+const WEIGHT_FAMILIES = ['tts', 'asr'];
 
 function readStoredFamily() {
   try {
@@ -55,27 +50,15 @@ function readStoredFamily() {
 
 export default function ModelCatalogue() {
   const { t } = useTranslation();
-  const pendingCatalogueTab = useAppStore((s) => s.pendingCatalogueTab);
   const pendingCatalogueFamily = useAppStore((s) => s.pendingCatalogueFamily);
-  const setPendingCatalogueTab = useAppStore((s) => s.setPendingCatalogueTab);
   const setPendingCatalogueFamily = useAppStore((s) => s.setPendingCatalogueFamily);
-  // Seed from the deep-link so the first paint is already the requested pane —
-  // seeding from storage and correcting in an effect would flash the wrong one.
-  const [pane, setPaneRaw] = useState(() =>
-    PANES.includes(pendingCatalogueTab) ? pendingCatalogueTab : readStoredPane(),
-  );
+  const openSettingsTab = useAppStore((s) => s.openSettingsTab);
+  // Seed from the deep-link so the first paint is already the requested
+  // family — seeding from storage and correcting in an effect would flash.
   const [family, setFamilyRaw] = useState(() =>
     FAMILIES.includes(pendingCatalogueFamily) ? pendingCatalogueFamily : readStoredFamily(),
   );
 
-  const setPane = useCallback((next) => {
-    setPaneRaw(next);
-    try {
-      localStorage.setItem(PANE_KEY, next);
-    } catch {
-      /* private mode / quota — the pane still switches, it just won't persist */
-    }
-  }, []);
   const setFamily = useCallback((next) => {
     setFamilyRaw(next);
     try {
@@ -85,53 +68,26 @@ export default function ModelCatalogue() {
     }
   }, []);
 
-  // Consume the one-shot deep-link (including a repeat request for the pane
-  // we're already on, which must still clear).
+  // Consume the one-shot deep-link (including a repeat request for the
+  // family we're already on, which must still clear).
   useEffect(() => {
-    if (!pendingCatalogueTab) return;
-    if (PANES.includes(pendingCatalogueTab)) setPane(pendingCatalogueTab);
+    if (!pendingCatalogueFamily) return;
     if (FAMILIES.includes(pendingCatalogueFamily)) setFamily(pendingCatalogueFamily);
-    setPendingCatalogueTab(null);
     setPendingCatalogueFamily(null);
-  }, [
-    pendingCatalogueTab,
-    pendingCatalogueFamily,
-    setPendingCatalogueFamily,
-    setPendingCatalogueTab,
-    setFamily,
-    setPane,
-  ]);
+  }, [pendingCatalogueFamily, setPendingCatalogueFamily, setFamily]);
 
-  const { data: info } = useSystemInfo();
-  const { data: status } = useModelStatus();
-
-  // The loaded-model pill ModelStoreTab renders in its header. Lives here now
-  // that this page — not Settings — hosts the model store.
-  const modelBadge =
-    status?.status === 'ready' ? (
-      <Badge tone="success">
-        <CheckCircle size={11} /> {t('models.ready_badge')}
-      </Badge>
-    ) : status?.status === 'loading' ? (
-      <Badge tone="warn">
-        <RefreshCw size={11} className="spinner" /> {t('models.loading_badge')}
-      </Badge>
-    ) : (
-      <Badge tone="warn">{t('models.idle_badge')}</Badge>
-    );
-
-  // Tabs, not a two-state Segmented: these are two workspaces of a catalogue,
-  // not one setting with an on/off reading, and the room to add a third pane
-  // later is free. Tabs also carry roving tabindex + role="tab" from the
-  // primitive, which the switch had to describe with an aria-label.
-  const paneItems = useMemo(
-    () => [
-      { id: 'engines', label: t('catalogue.tab_engines'), icon: Cpu },
-      { id: 'models', label: t('catalogue.tab_models'), icon: HardDriveDownload },
-    ],
-    [t],
+  const enginesRef = useRef(null);
+  const changeFamily = useCallback(
+    (next) => {
+      setFamily(next);
+      enginesRef.current?.scrollIntoView?.({ behavior: 'smooth', block: 'start' });
+    },
+    [setFamily],
   );
 
+  const { data: models } = useModels();
+  const cacheDir = models?.hf_cache_dir?.replace(/^\/Users\/[^/]+/, '~');
+
   return (
     // Same container-query shell as Settings: the app is zoom-scaled, so
     // viewport media queries fire at the wrong logical width under Tauri.
@@ -140,48 +96,83 @@ export default function ModelCatalogue() {
       data-testid="model-catalogue"
     >
       <div className="mx-auto box-border w-full max-w-[1500px] px-[34px] pb-[56px] pt-[34px] font-sans @max-[900px]/catalogue-shell:px-[24px] @max-[900px]/catalogue-shell:pb-[36px] @max-[900px]/catalogue-shell:pt-[26px] @max-[560px]/catalogue-shell:px-[14px]">
-        {/* Treat this as a workspace, not a Settings card: one editorial title,
-            one quiet navigation line, then the data surface. */}
-        <header className="mb-[24px] flex flex-wrap items-end justify-between gap-x-[32px] gap-y-[18px] border-b border-[color-mix(in_srgb,var(--chrome-fg)_9%,transparent)] pb-[18px] @max-[560px]/catalogue-shell:flex-col @max-[560px]/catalogue-shell:items-start">
-          <div className="flex items-center gap-[12px]">
-            <span
-              className="inline-flex h-[34px] w-[34px] shrink-0 items-center justify-center rounded-[12px] bg-[color-mix(in_srgb,var(--chrome-accent)_11%,transparent)] text-[color:var(--chrome-accent)]"
-              aria-hidden="true"
-            >
-              <Boxes size={17} strokeWidth={1.6} />
-            </span>
-            <h1 className="m-0 min-w-0 [font-family:var(--font-serif)] text-[2rem] font-normal leading-none tracking-[-0.025em] text-[color:var(--chrome-fg)] @max-[560px]/catalogue-shell:text-[1.55rem]">
-              {t('catalogue.title')}
-            </h1>
-          </div>
-          <Tabs
-            items={paneItems}
-            value={pane}
-            onChange={setPane}
-            variant="underline"
-            idPrefix="catalogue-pane"
-            aria-label={t('catalogue.title')}
-            data-testid="catalogue-pane-switch"
-          />
+        <header className="mb-[22px] flex items-center gap-[12px]">
+          <span
+            className="inline-flex h-[34px] w-[34px] shrink-0 items-center justify-center rounded-[12px] bg-[color-mix(in_srgb,var(--chrome-accent)_11%,transparent)] text-[color:var(--chrome-accent)]"
+            aria-hidden="true"
+          >
+            <Boxes size={17} strokeWidth={1.6} />
+          </span>
+          <h1 className="m-0 min-w-0 [font-family:var(--font-serif)] text-[2rem] font-normal leading-none tracking-[-0.025em] text-[color:var(--chrome-fg)] @max-[560px]/catalogue-shell:text-[1.55rem]">
+            {t('catalogue.title')}
+          </h1>
         </header>
 
-        {/* Full-bleed: engine rows and the model table are wide, data-dense
+        <SetupSummary onChange={changeFamily} />
+
+        {/* Full-bleed: engine rows and the weights table are wide, data-dense
             surfaces — a reading-width column would only add horizontal
             scrolling inside them. */}
-        <div
-          key={pane}
-          id={`catalogue-pane-panel-${pane}`}
-          data-testid={`catalogue-pane-${pane}`}
-          role="tabpanel"
-          aria-labelledby={`catalogue-pane-tab-${pane}`}
-          className="min-w-0 [&>*:first-child]:mt-0"
+        <section
+          ref={enginesRef}
+          data-testid="catalogue-engines"
+          className="min-w-0 scroll-mt-[24px] [&>*:first-child]:mt-0"
         >
-          {pane === 'engines' ? (
-            <EnginesTab initialFamily={family} onFamilyChange={setFamily} catalogueLayout />
-          ) : (
-            <ModelStoreTab info={info} modelBadge={modelBadge} catalogueLayout />
+          <EnginesTab initialFamily={family} onFamilyChange={setFamily} catalogueLayout />
+        </section>
+
+        {WEIGHT_FAMILIES.includes(family) && (
+          <section
+            data-testid="catalogue-weights"
+            aria-labelledby="catalogue-weights-title"
+            className="mt-[40px] min-w-0"
+          >
+            <h2
+              id="catalogue-weights-title"
+              className="m-0 mb-[14px] px-[2px] text-[length:var(--text-lg)] font-semibold text-[color:var(--chrome-fg)]"
+            >
+              {t('catalogue.weights')}
+            </h2>
+            <ModelStoreTab key={family} family={family} />
+          </section>
+        )}
+
+        <footer
+          data-testid="catalogue-storage"
+          className="mt-[40px] flex flex-wrap items-center gap-x-[10px] gap-y-[4px] border-t border-border pt-[14px] font-mono text-[11px] text-muted-foreground"
+        >
+          {models && (
+            <>
+              <span className="text-foreground">
+                {fmtBytes(models.total_installed_bytes) || '0 B'}
+              </span>
+              {models.disk_free_gb != null && (
+                <>
+                  <span aria-hidden="true">·</span>
+                  <span title={t('models.disk_free_title')}>
+                    {t('models.disk_free', { size: `${models.disk_free_gb} GB` })}
+                  </span>
+                </>
+              )}
+              {cacheDir && (
+                <>
+                  <span aria-hidden="true">·</span>
+                  <code className="truncate" title={models.hf_cache_dir}>
+                    {cacheDir}
+                  </code>
+                </>
+              )}
+            </>
           )}
-        </div>
+          <button
+            type="button"
+            className="ml-auto cursor-pointer border-0 bg-transparent p-0 font-mono text-[11px] text-accent hover:underline"
+            onClick={() => openSettingsTab('storage')}
+            data-testid="catalogue-storage-link"
+          >
+            {t('catalogue.storage_link')} →
+          </button>
+        </footer>
       </div>
     </div>
   );

frontend/src/pages/ModelCatalogue.test.jsx

@@ -1,160 +1,151 @@
-// Model Catalogue workspace — pane switching, deep-link hand-off, persistence.
+// Model Catalogue workspace — one family axis, deep-link hand-off, persistence.
 //
-// The two panes are mounted stand-ins for the real EnginesTab / ModelStoreTab
-// (both of which have their own suites and both of which hit the network); this
-// suite is about the workspace shell around them.
+// The summary, engine list and weights list are mounted stand-ins for the real
+// components (each has its own suite and hits the network); this suite is
+// about the page around them: one family drives both lists, LLM has no
+// weights, Change on the summary switches the family, and the old two-pane
+// switch is gone.
 import { describe, it, expect, beforeEach, vi } from 'vitest';
 import React from 'react';
 import { render, screen, fireEvent, act } from '@testing-library/react';
 
 vi.mock('../components/settings/EnginesTab', () => ({
-  default: ({ initialFamily, catalogueLayout }) => (
+  default: ({ initialFamily, onFamilyChange, catalogueLayout }) => (
     <div data-testid="stub-engines" data-catalogue-layout={catalogueLayout || undefined}>
       {initialFamily}
+      <button type="button" onClick={() => onFamilyChange('llm')}>
+        to-llm
+      </button>
     </div>
   ),
 }));
 vi.mock('../components/settings/ModelStoreTab', () => ({
-  default: ({ modelBadge, catalogueLayout }) => (
-    <div data-testid="stub-models" data-catalogue-layout={catalogueLayout || undefined}>
-      {modelBadge}
+  default: ({ family }) => <div data-testid="stub-weights">{family}</div>,
+}));
+vi.mock('../components/catalogue/SetupSummary', () => ({
+  default: ({ onChange }) => (
+    <div data-testid="stub-summary">
+      <button type="button" onClick={() => onChange('asr')}>
+        change-asr
+      </button>
     </div>
   ),
 }));
 vi.mock('../api/hooks', () => ({
-  useSystemInfo: () => ({ data: { has_hf_token: false } }),
-  useModelStatus: () => ({ data: { status: 'ready' } }),
+  useModels: () => ({
+    data: {
+      total_installed_bytes: 3 * 1024 ** 3,
+      disk_free_gb: 41,
+      hf_cache_dir: '/Users/dev/.cache/huggingface',
+    },
+  }),
 }));
 
 import { useAppStore } from '../store';
 import ModelCatalogue from './ModelCatalogue';
 
-// Radix Tabs activate on POINTER DOWN, not click — a bare fireEvent.click
-// leaves the pane unchanged and reads as a broken switcher. Drive the tab the
-// way a pointer does.
-const clickTab = (name) => {
-  const tab = screen.getByRole('tab', { name });
-  fireEvent.pointerDown(tab, { button: 0, ctrlKey: false, pointerType: 'mouse' });
-  fireEvent.mouseDown(tab, { button: 0 });
-  fireEvent.click(tab);
-};
-
 describe('ModelCatalogue', () => {
   beforeEach(() => {
     localStorage.clear();
     act(() => {
-      useAppStore.setState({
-        mode: 'catalogue',
-        pendingCatalogueTab: null,
-        pendingCatalogueFamily: null,
-      });
+      useAppStore.setState({ mode: 'catalogue', pendingCatalogueFamily: null });
     });
   });
 
-  it('opens on the Engines pane and switches to Models', () => {
+  it('is one page: summary, then the TTS engines and TTS weights, no pane switch', () => {
     render(<ModelCatalogue />);
-    expect(screen.getByTestId('stub-engines')).toBeInTheDocument();
-    expect(screen.getByRole('tabpanel', { name: 'Engines' })).toBeInTheDocument();
+    expect(screen.getByTestId('stub-summary')).toBeInTheDocument();
+    expect(screen.getByTestId('stub-engines')).toHaveTextContent('tts');
     expect(screen.getByTestId('stub-engines')).toHaveAttribute('data-catalogue-layout', 'true');
-    expect(screen.queryByTestId('stub-models')).toBeNull();
-
-    clickTab('Models');
-    expect(screen.getByTestId('stub-models')).toBeInTheDocument();
-    expect(screen.getByRole('tabpanel', { name: 'Models' })).toBeInTheDocument();
-    expect(screen.getByTestId('stub-models')).toHaveAttribute('data-catalogue-layout', 'true');
-    expect(screen.queryByTestId('stub-engines')).toBeNull();
+    expect(screen.getByTestId('stub-weights')).toHaveTextContent('tts');
+    expect(screen.queryByRole('tab')).toBeNull();
+    // Reading order: summary above engines above weights.
+    const [summary, engines, weights] = ['stub-summary', 'stub-engines', 'stub-weights'].map((id) =>
+      screen.getByTestId(id),
+    );
+    expect(
+      summary.compareDocumentPosition(engines) & Node.DOCUMENT_POSITION_FOLLOWING,
+    ).toBeTruthy();
+    expect(
+      engines.compareDocumentPosition(weights) & Node.DOCUMENT_POSITION_FOLLOWING,
+    ).toBeTruthy();
   });
 
-  it('uses the wide workspace shell for data-heavy catalogue panes', () => {
+  it('Change on the summary switches the family for engines and weights together', () => {
+    render(<ModelCatalogue />);
+    fireEvent.click(screen.getByText('change-asr'));
+    expect(screen.getByTestId('stub-engines')).toHaveTextContent('asr');
+    expect(screen.getByTestId('stub-weights')).toHaveTextContent('asr');
+  });
+
+  it('LLM engines bring their own weights, so the weights block disappears', () => {
+    render(<ModelCatalogue />);
+    fireEvent.click(screen.getByText('to-llm'));
+    expect(screen.getByTestId('stub-engines')).toHaveTextContent('llm');
+    expect(screen.queryByTestId('stub-weights')).toBeNull();
+  });
+
+  it('uses the wide workspace shell for data-heavy lists', () => {
     render(<ModelCatalogue />);
     expect(screen.getByTestId('model-catalogue').firstElementChild).toHaveClass('max-w-[1500px]');
   });
 
-  it('associates each tab with its labelled panel', () => {
+  it('remembers the family across visits', () => {
+    const { unmount } = render(<ModelCatalogue />);
+    fireEvent.click(screen.getByText('change-asr'));
+    unmount();
     render(<ModelCatalogue />);
-    const enginesTab = screen.getByRole('tab', { name: 'Engines' });
-    const enginesPanel = screen.getByRole('tabpanel', { name: 'Engines' });
-    expect(enginesTab).toHaveAttribute('aria-controls', enginesPanel.id);
-    expect(enginesPanel).toHaveAttribute('aria-labelledby', enginesTab.id);
-
-    const modelsTab = screen.getByRole('tab', { name: 'Models' });
-    clickTab('Models');
-    const modelsPanel = screen.getByRole('tabpanel', { name: 'Models' });
-    expect(modelsTab).toHaveAttribute('aria-controls', modelsPanel.id);
-    expect(modelsPanel).toHaveAttribute('aria-labelledby', modelsTab.id);
+    expect(screen.getByTestId('stub-engines')).toHaveTextContent('asr');
   });
 
-  it('remembers the pane across visits', () => {
-    const first = render(<ModelCatalogue />);
-    clickTab('Models');
-    first.unmount();
-
+  it('honours a pending family deep-link on first paint and clears it', () => {
+    act(() => useAppStore.setState({ pendingCatalogueFamily: 'llm' }));
     render(<ModelCatalogue />);
-    expect(screen.getByTestId('stub-models')).toBeInTheDocument();
+    expect(screen.getByTestId('stub-engines')).toHaveTextContent('llm');
+    expect(useAppStore.getState().pendingCatalogueFamily).toBeNull();
   });
 
-  it('honours a pending deep-link pane on first paint and clears it', () => {
+  it('switches family when a deep-link arrives while already mounted', () => {
+    render(<ModelCatalogue />);
+    expect(screen.getByTestId('stub-engines')).toHaveTextContent('tts');
     act(() => {
-      useAppStore.setState({ pendingCatalogueTab: 'models' });
+      useAppStore.getState().openCatalogue('asr');
     });
-    render(<ModelCatalogue />);
-    expect(screen.getByTestId('stub-models')).toBeInTheDocument();
-    expect(useAppStore.getState().pendingCatalogueTab).toBeNull();
-  });
-
-  it('switches pane when a deep-link arrives while already mounted', () => {
-    render(<ModelCatalogue />);
-    expect(screen.getByTestId('stub-engines')).toBeInTheDocument();
-
-    act(() => {
-      useAppStore.getState().openCatalogue('models');
-    });
-    expect(screen.getByTestId('stub-models')).toBeInTheDocument();
-    expect(useAppStore.getState().pendingCatalogueTab).toBeNull();
-  });
-
-  it('honours an engine-family deep link and clears it after hand-off', () => {
-    act(() => {
-      useAppStore.getState().openCatalogue({ pane: 'engines', family: 'asr' });
-    });
-    render(<ModelCatalogue />);
     expect(screen.getByTestId('stub-engines')).toHaveTextContent('asr');
     expect(useAppStore.getState().pendingCatalogueFamily).toBeNull();
   });
 
-  it('passes the loaded-model badge down to the model store pane', () => {
-    act(() => {
-      useAppStore.setState({ pendingCatalogueTab: 'models' });
-    });
+  it('shows one storage line and links it to Settings → Storage', () => {
     render(<ModelCatalogue />);
-    // 'ready' status → the ready badge renders inside the models pane.
-    expect(screen.getByTestId('stub-models').textContent).toBeTruthy();
+    const footer = screen.getByTestId('catalogue-storage');
+    expect(footer).toHaveTextContent('41 GB');
+    expect(footer).toHaveTextContent('~/.cache/huggingface');
+    fireEvent.click(screen.getByTestId('catalogue-storage-link'));
+    expect(useAppStore.getState().mode).toBe('settings');
+    expect(useAppStore.getState().pendingSettingsTab).toBe('storage');
   });
 });
 
 describe('openCatalogue', () => {
-  it('navigates to the catalogue workspace on the requested pane', () => {
-    act(() => {
-      useAppStore.setState({ mode: 'settings', pendingCatalogueTab: null });
-      useAppStore.getState().openCatalogue('models');
-    });
+  beforeEach(() => {
+    act(() => useAppStore.setState({ mode: 'studio', pendingCatalogueFamily: null }));
+  });
+
+  it('navigates to the catalogue on the requested family', () => {
+    act(() => useAppStore.getState().openCatalogue('llm'));
     expect(useAppStore.getState().mode).toBe('catalogue');
-    expect(useAppStore.getState().pendingCatalogueTab).toBe('models');
-  });
-
-  it('defaults to the engines pane', () => {
-    act(() => {
-      useAppStore.setState({ pendingCatalogueTab: null });
-      useAppStore.getState().openCatalogue();
-    });
-    expect(useAppStore.getState().pendingCatalogueTab).toBe('engines');
-  });
-
-  it('accepts an object deep link with a family', () => {
-    act(() => {
-      useAppStore.getState().openCatalogue({ pane: 'engines', family: 'llm' });
-    });
-    expect(useAppStore.getState().pendingCatalogueTab).toBe('engines');
     expect(useAppStore.getState().pendingCatalogueFamily).toBe('llm');
   });
+
+  it('accepts the object form and ignores the retired pane key', () => {
+    act(() => useAppStore.getState().openCatalogue({ pane: 'models', family: 'asr' }));
+    expect(useAppStore.getState().mode).toBe('catalogue');
+    expect(useAppStore.getState().pendingCatalogueFamily).toBe('asr');
+  });
+
+  it('defaults to the last family when called bare', () => {
+    act(() => useAppStore.getState().openCatalogue());
+    expect(useAppStore.getState().mode).toBe('catalogue');
+    expect(useAppStore.getState().pendingCatalogueFamily).toBeNull();
+  });
 });

frontend/src/pages/Settings.jsx

@@ -35,7 +35,7 @@ import UpdatesPanel from '../components/UpdatesPanel';
 import GeneralTab from '../components/settings/GeneralTab';
 // Engine selection + the model store moved to the Model Catalogue workspace
 // (pages/ModelCatalogue.jsx); these categories now signpost it.
-import CataloguePointer from '../components/settings/CataloguePointer';
+import VoicePreviewsPanel from '../components/settings/VoicePreviewsPanel';
 import HotkeyTab from '../components/settings/HotkeyTab';
 import TranslationTab from '../components/settings/TranslationTab';
 import NetworkTab from '../components/settings/NetworkTab';
@@ -399,20 +399,6 @@ export default function Settings() {
             <GeneralTab />
           </>
         );
-      case 'engines':
-        return <CataloguePointer area="engines" />;
-      case 'models':
-        // What stays here is the storage-shaped remainder: where weights live
-        // on disk and which mirror they come from (both restart-bound, hence
-        // the category's `restart: true`). Browsing / installing / removing
-        // them is the Model Catalogue's Models pane.
-        return (
-          <>
-            <StoragePanel />
-            <HFMirrorPanel />
-            <CataloguePointer area="models" />
-          </>
-        );
       case 'dictation':
         return (
           <>
@@ -436,16 +422,28 @@ export default function Settings() {
           </>
         );
       case 'storage':
+        // Everything disk-shaped in one place: usage, the models directory
+        // (restart-bound), data/output paths, and the voice-preview cache
+        // toggle. Browsing / installing / removing weights is the Model
+        // Catalogue.
         return (
           <>
             <StorageUsagePanel />
+            <StoragePanel />
             <StorageTab />
+            <VoicePreviewsPanel />
           </>
         );
       case 'permissions':
         return <PermissionsPanel />;
       case 'network':
-        return <NetworkTab />;
+        // The HF mirror is a network choice (which host downloads come from).
+        return (
+          <>
+            <NetworkTab />
+            <HFMirrorPanel />
+          </>
+        );
       case 'audio-tools':
         return <AudioToolsPanel />;
       case 'sharing':

frontend/src/store/uiSlice.ts

@@ -33,8 +33,12 @@ export type AppMode =
   | 'settings';
 
 /** Which pane the Model Catalogue workspace opens on. */
-export type CatalogueTab = 'engines' | 'models';
-export type CatalogueTarget = CatalogueTab | { pane?: CatalogueTab; family?: EngineFamily };
+/**
+ * Deep-link target for the Model Catalogue: the engine family to open on. The
+ * page has one axis now (TTS / ASR / LLM) — the old `pane` key is tolerated
+ * and ignored so stale callers keep navigating.
+ */
+export type CatalogueTarget = EngineFamily | { family?: EngineFamily | null; pane?: string };
 
 /**
  * The Voice workspace's "Define voice" method (was the Clone/Design tab
@@ -77,12 +81,9 @@ export interface UiSlice {
    */
   pendingSettingsTab: string | null;
   /**
-   * One-shot hand-off for "open the Model Catalogue on a specific pane" — the
-   * catalogue twin of `pendingSettingsTab`, used by the Settings pointers that
-   * replaced the old Engines / Model Store panels.
+   * One-shot hand-off for "open the Model Catalogue on a specific engine
+   * family" — the catalogue twin of `pendingSettingsTab`.
    */
-  pendingCatalogueTab: CatalogueTab | null;
-  /** Optional engine family to focus after entering the catalogue. */
   pendingCatalogueFamily: EngineFamily | null;
   isSidebarCollapsed: boolean;
   isSidebarProjectsCollapsed: boolean;
@@ -102,12 +103,11 @@ export interface UiSlice {
   setModeBeforeVoice: (mode: AppMode | null) => void;
   setPendingProfileId: (id: string | null) => void;
   setPendingSettingsTab: (tab: string | null) => void;
-  setPendingCatalogueTab: (tab: CatalogueTab | null) => void;
   setPendingCatalogueFamily: (family: EngineFamily | null) => void;
   /** Navigate to Settings on a specific tab in one call. */
   openSettingsTab: (tab: string) => void;
-  /** Navigate to the Model Catalogue on a specific pane in one call. */
-  openCatalogue: (target?: CatalogueTarget) => void;
+  /** Navigate to the Model Catalogue on a specific engine family in one call. */
+  openCatalogue: (target?: CatalogueTarget | null) => void;
   setIsSidebarCollapsed: (collapsed: boolean) => void;
   setIsSidebarProjectsCollapsed: (collapsed: boolean) => void;
   setSidebarTab: (tab: SidebarTab) => void;
@@ -132,7 +132,6 @@ export const createUiSlice: StateCreator<UiSlice, [], [], UiSlice> = (set, get)
   modeBeforeVoice: null,
   pendingProfileId: null,
   pendingSettingsTab: null,
-  pendingCatalogueTab: null,
   pendingCatalogueFamily: null,
   isSidebarCollapsed: false,
   isSidebarProjectsCollapsed: false,
@@ -154,13 +153,11 @@ export const createUiSlice: StateCreator<UiSlice, [], [], UiSlice> = (set, get)
   setModeBeforeVoice: (mode) => set({ modeBeforeVoice: mode }),
   setPendingProfileId: (id) => set({ pendingProfileId: id }),
   setPendingSettingsTab: (tab) => set({ pendingSettingsTab: tab }),
-  setPendingCatalogueTab: (tab) => set({ pendingCatalogueTab: tab }),
   setPendingCatalogueFamily: (family) => set({ pendingCatalogueFamily: family }),
   openSettingsTab: (tab) => set({ pendingSettingsTab: tab, mode: 'settings' }),
-  openCatalogue: (target = 'engines') => {
-    const { pane = 'engines', family = null } =
-      typeof target === 'string' ? { pane: target } : target;
-    set({ pendingCatalogueTab: pane, pendingCatalogueFamily: family, mode: 'catalogue' });
+  openCatalogue: (target = null) => {
+    const family = typeof target === 'string' ? target : (target?.family ?? null);
+    set({ pendingCatalogueFamily: family, mode: 'catalogue' });
   },
   setIsSidebarCollapsed: (collapsed) => set({ isSidebarCollapsed: collapsed }),
   setIsSidebarProjectsCollapsed: (collapsed) => set({ isSidebarProjectsCollapsed: collapsed }),

frontend/src/test/EngineCompatibilityMatrix.test.jsx

@@ -1153,7 +1153,7 @@ describe('EngineCompatibilityMatrix', () => {
       />,
     );
     await waitFor(() => screen.getByText('OmniVoice (test)'));
-    expect(screen.getByText('Engine Compatibility Matrix')).toBeInTheDocument();
+    expect(screen.getByRole('heading', { name: 'Engines' })).toBeInTheDocument();
     expect(document.querySelectorAll('.engine-matrix__tab-family').length).toBe(3);
     const labels = document.querySelectorAll('.engine-matrix__tab-label');
     expect(labels).toHaveLength(3);

frontend/src/test/modelStoreGrouping.test.jsx

@@ -228,8 +228,9 @@ vi.mock('../components/settings/models/ModelsTable', () => ({
 }));
 
 import ModelStoreTab from '../components/settings/ModelStoreTab';
+import { scopeReco } from '../components/settings/models/sections';
 
-function mountTab() {
+function mountTab(props = {}) {
   global.EventSource = class {
     constructor() {
       this.onmessage = null;
@@ -238,7 +239,7 @@ function mountTab() {
   };
   return render(
     <I18nextProvider i18n={i18n}>
-      <ModelStoreTab info={{ has_hf_token: true }} modelBadge={null} />
+      <ModelStoreTab info={{ has_hf_token: true }} {...props} />
     </I18nextProvider>,
   );
 }
@@ -251,7 +252,7 @@ describe('Model Store — grouped catalog', () => {
   it('renders role sections in order with localized titles', async () => {
     mountTab();
     await waitFor(() => expect(screen.getByTestId('models-section-tts')).toBeInTheDocument());
-    expect(screen.getByTestId('model-list-panel')).toHaveClass('min-h-[95%]');
+    expect(screen.getByTestId('model-list-panel')).toBeInTheDocument();
     expect(screen.getByTestId('model-list-area')).toHaveClass('flex-1');
     const keys = ['tts', 'asr', 'dictation', 'diarisation'];
     const sections = keys.map((k) => screen.getByTestId(`models-section-${k}`));
@@ -313,3 +314,50 @@ describe('Model Store — grouped catalog', () => {
     expect(search).toHaveValue('');
   });
 });
+
+// ── Family scoping: the catalogue lists one family's weights under its engines ─
+
+describe('Model Store — scoped to one engine family', () => {
+  it('ASR carries offline ASR, streaming dictation and diarisation; TTS only TTS', async () => {
+    mountTab({ family: 'asr' });
+    await waitFor(() => expect(screen.getByTestId('models-section-asr')).toBeInTheDocument());
+    expect(screen.getByTestId('models-section-dictation')).toBeInTheDocument();
+    expect(screen.getByTestId('models-section-diarisation')).toBeInTheDocument();
+    expect(screen.queryByTestId('models-section-tts')).not.toBeInTheDocument();
+    expect(visibleLabels()).not.toContain('VoiceStudio TTS');
+  });
+
+  it('TTS shows only TTS weights and searching within it stays scoped', async () => {
+    mountTab({ family: 'tts' });
+    await waitFor(() => expect(screen.getByTestId('models-section-tts')).toBeInTheDocument());
+    expect(screen.queryByTestId('models-section-asr')).not.toBeInTheDocument();
+    const search = screen.getByRole('searchbox', { name: t('models.search_label') });
+    fireEvent.change(search, { target: { value: 'parakeet' } });
+    // Parakeet is an ASR row — no match inside TTS, so the scoped empty state shows.
+    await waitFor(() => expect(screen.getByText(t('models.no_matches'))).toBeInTheDocument());
+  });
+
+  it('scopeReco narrows the device preset to the family and recomputes totals', () => {
+    const reco = {
+      device: { label: 'M2' },
+      all_installed: false,
+      download_gb_remaining: 5.3,
+      total_gb: 7.7,
+      models: [
+        { repo_id: 'a/tts', role: 'TTS', installed: true, size_gb: 2.4 },
+        { repo_id: 'b/asr', role: 'ASR', installed: false, size_gb: 2.9 },
+        { repo_id: 'c/dict', role: 'ASR', engine: 'sherpa-onnx', installed: false, size_gb: 2.4 },
+      ],
+    };
+    expect(scopeReco(reco, null)).toBe(reco);
+    const tts = scopeReco(reco, 'tts');
+    expect(tts.models.map((m) => m.repo_id)).toEqual(['a/tts']);
+    expect(tts.all_installed).toBe(true);
+    expect(tts.download_gb_remaining).toBe(0);
+    const asr = scopeReco(reco, 'asr');
+    expect(asr.models.map((m) => m.repo_id)).toEqual(['b/asr', 'c/dict']);
+    expect(asr.all_installed).toBe(false);
+    expect(asr.download_gb_remaining).toBe(5.3);
+    expect(scopeReco(reco, 'llm')).toBeNull();
+  });
+});

frontend/src/test/modelStoreTokenPath.test.jsx

@@ -1,88 +0,0 @@
-import React from 'react';
-import { describe, it, expect, vi, beforeEach, afterEach } from 'vitest';
-import { render, screen, fireEvent, waitFor } from '@testing-library/react';
-import { I18nextProvider } from 'react-i18next';
-import i18n from '../i18n';
-
-// ── Enhancement: unify the two HF-token entry points ───────────────────────
-// The Models toolbar used to save the token via /system/set-env (env var +
-// HF-CLI file) while Settings → Credentials saves to the encrypted app store —
-// two stores with an asymmetric clear path (Credentials' "Clear" couldn't
-// remove a toolbar-set token). This test locks the toolbar onto the SAME
-// canonical app-store endpoint Credentials uses (/api/settings/hf-token), and
-// guards against a regression back to /system/set-env.
-
-const apiPost = vi.fn(() => Promise.resolve({ active: 'app', sources: [] }));
-vi.mock('../api/client', () => ({
-  apiPost: (...a) => apiPost(...a),
-  apiFetch: vi.fn(() => Promise.resolve({ ok: true, json: () => Promise.resolve({}) })),
-}));
-
-const refetch = vi.fn();
-vi.mock('../api/hooks', () => ({
-  useModels: () => ({
-    data: {
-      models: [],
-      total_installed_bytes: 0,
-      disk_free_gb: 42.5,
-      hf_cache_dir: '/home/u/.cache/huggingface',
-    },
-    isLoading: false,
-    refetch,
-  }),
-  useRecommendations: () => ({ data: null, refetch }),
-  useInstallModel: () => ({ mutateAsync: vi.fn() }),
-  useDeleteModel: () => ({ mutateAsync: vi.fn() }),
-}));
-
-vi.mock('../api/setup', () => ({
-  setupDownloadStreamUrl: () => 'http://localhost/stream',
-  cancelInstallModel: vi.fn(),
-}));
-vi.mock('../api/external', () => ({ openExternal: vi.fn() }));
-// Keep the render light + focused on the toolbar.
-vi.mock('../components/settings/models/ModelsTable', () => ({ default: () => null }));
-vi.mock('../components/settings/models/RecoBanner', () => ({ default: () => null }));
-
-import ModelStoreTab from '../components/settings/ModelStoreTab';
-
-function withI18n(node) {
-  return <I18nextProvider i18n={i18n}>{node}</I18nextProvider>;
-}
-
-describe('Model Store toolbar — HF token saves to the canonical app store', () => {
-  beforeEach(() => {
-    apiPost.mockClear();
-    global.EventSource = class {
-      constructor() {
-        this.onmessage = null;
-      }
-      close() {}
-    };
-  });
-  afterEach(() => vi.restoreAllMocks());
-
-  async function openAndSave(token) {
-    render(withI18n(<ModelStoreTab info={{ has_hf_token: false }} modelBadge={null} />));
-    // The compact toolbar shows a "HF Token" trigger when no token is set.
-    fireEvent.click(screen.getByRole('button', { name: /HF Token/i }));
-    const input = screen.getByPlaceholderText(/hf_/i);
-    fireEvent.change(input, { target: { value: token } });
-    fireEvent.click(screen.getByRole('button', { name: i18n.t('common.save') }));
-  }
-
-  it('POSTs the token to /api/settings/hf-token (same store as Credentials)', async () => {
-    await openAndSave('hf_abc123');
-    await waitFor(() => expect(apiPost).toHaveBeenCalledTimes(1));
-    expect(apiPost).toHaveBeenCalledWith('/api/settings/hf-token', { token: 'hf_abc123' });
-  });
-
-  it('never routes the token through the legacy /system/set-env store', async () => {
-    await openAndSave('hf_xyz');
-    await waitFor(() => expect(apiPost).toHaveBeenCalled());
-    for (const call of apiPost.mock.calls) {
-      expect(call[0]).not.toContain('/system/set-env');
-    }
-    expect(apiPost).toHaveBeenCalledWith('/api/settings/hf-token', { token: 'hf_xyz' });
-  });
-});