Show model download progress on first start

README.md

@@ -97,6 +97,8 @@ This will:
 3. Summarize with your local LLM
 4. Save everything to `~/ownscribe/YYYY-MM-DD_HHMMSS/`
 
+On first run, WhisperX / pyannote may download model files. ownscribe shows a `Preparing models` step and best-effort download progress in the TUI while this happens.
+
 ### Options
 
 ```bash
@@ -117,6 +119,7 @@ ownscribe --template lecture                  # use the lecture summarization te
 ```bash
 ownscribe devices                  # list audio devices (uses native CoreAudio when available)
 ownscribe apps                     # list running apps with PIDs for use with --pid
+ownscribe warmup                   # prefetch WhisperX/pyannote models before a meeting
 ownscribe transcribe recording.wav # transcribe an audio file (saves alongside the input)
 ownscribe summarize transcript.md  # summarize a transcript (saves alongside the input)
 ownscribe resume ./2026-02-20_1736 # resume a failed/partial pipeline in a directory
@@ -125,6 +128,14 @@ ownscribe config                   # open config file in $EDITOR
 ownscribe cleanup                  # remove ownscribe data from disk
 ```
 
+Use `warmup` ahead of time to avoid first-run model download delays while recording:
+
+```bash
+ownscribe warmup                    # prefetch Whisper model (+ diarization if enabled in config)
+ownscribe warmup --language en      # also prefetch alignment model for English
+ownscribe warmup --with-diarization # force diarization warmup for this run
+```
+
 ### Searching Meeting Notes
 
 Use `ask` to search across all your meeting notes with natural-language questions:

src/ownscribe/cli.py

@@ -149,6 +149,35 @@ def transcribe(
     run_transcribe(config, file)
 
 
+@cli.command()
+@click.option("--model", default=None, help="Whisper model size.")
+@click.option("--language", default=None, help="Language code to prefetch alignment model for (e.g. en, de, fr).")
+@click.option(
+    "--with-diarization/--no-diarization",
+    "with_diarization",
+    default=None,
+    help="Override diarization warmup (defaults to config setting).",
+)
+@click.pass_context
+def warmup(
+    ctx: click.Context,
+    model: str | None,
+    language: str | None,
+    with_diarization: bool | None,
+) -> None:
+    """Prefetch WhisperX/pyannote models to avoid first-run stalls."""
+    config = ctx.obj["config"]
+    if model:
+        config.transcription.model = model
+    if language:
+        config.transcription.language = language
+    if with_diarization is not None:
+        config.diarization.enabled = with_diarization
+
+    from ownscribe.pipeline import run_warmup
+    run_warmup(config)
+
+
 @cli.command()
 @click.argument("file", type=click.Path(exists=True))
 @click.option("--template", default=None, help="Summarization template (meeting, lecture, brief, or custom).")

src/ownscribe/pipeline.py

@@ -222,6 +222,42 @@ def run_transcribe(config: Config, audio_file: str) -> None:
     _do_transcribe_and_summarize(config, audio_path, out_dir, summarize=False)
 
 
+def run_warmup(config: Config) -> None:
+    """Prefetch transcription/diarization models without processing audio."""
+    diar_enabled = config.diarization.enabled and bool(config.diarization.hf_token)
+    hf_token_warning = (
+        config.diarization.enabled and not config.diarization.hf_token
+    )
+
+    with PipelineProgress(diarize=False, summarize=False, transcribe=False, include_prepare=True) as progress:
+        try:
+            transcriber = _create_transcriber(config, progress=progress)
+        except ImportError:
+            click.echo(
+                "Error: WhisperX is not installed. Install with:\n"
+                "  uv pip install 'ownscribe[transcription]'",
+                err=True,
+            )
+            raise SystemExit(1) from None
+
+        transcriber.prepare_models(language=config.transcription.language or None)
+
+    click.echo(f"Whisper model ready: {config.transcription.model}")
+    if config.transcription.language:
+        click.echo(f"Alignment model ready: {config.transcription.language}")
+    else:
+        click.echo("Alignment model not preloaded (language auto-detect).")
+
+    if diar_enabled:
+        click.echo("Diarization pipeline ready.")
+    elif hf_token_warning:
+        click.echo(
+            "Warning: Diarization enabled but no HF token configured. "
+            "Skipping diarization warmup.",
+            err=True,
+        )
+
+
 def run_summarize(config: Config, transcript_file: str) -> None:
     """Summarize a transcript file and save the summary alongside the input."""
     transcript_path = Path(transcript_file).resolve()