Version: 1.0.8
No file selected.
Supported formats: Any video or audio format supported by ffmpeg — including MP4, MOV, MKV, WebM, AVI, MPEG, MP3, WAV, M4A, AAC, OGG, FLAC, Opus, WMA.
No active jobs.
Returns text/event-stream. JSON body with gcs_url or multipart upload at POST /transcribe/upload.
{
"gcs_url": "https://storage.googleapis.com/bucket/video.mp4",
"model": "small",
"hotwords": ["AdminPulse", "AccountancyVandaag"]
}
Pass model to select a whisper model. Multilingual models auto-detect language; language-specific entries load fine-tuned weights where available or pin a multilingual model to a language (large-v3.nl, large-v3.fr). Legacy quality presets still work (fast → small, balanced → medium, best → large-v3-turbo).
| model | languages | weights | compute | RAM / job |
|---|---|---|---|---|
small | multilingual | small | int8 | ~2 GB |
medium | multilingual | medium | int8 | ~4 GB |
large-v3-turbo | multilingual | large-v3-turbo | int8 | ~7 GB |
distil-large-v3 | multilingual | distil-large-v3 | int8 | ~9 GB |
large-v3 | multilingual | large-v3 | int8 | ~14 GB |
small.en | en | small.en | int8 | ~2 GB |
medium.en | en | medium.en | int8 | ~3 GB |
distil-small.en | en | distil-small.en | int8 | ~2 GB |
distil-medium.en | en | distil-medium.en | int8 | ~3 GB |
small.nl | nl | ammaraldirawi/faster-whisper-small-nl-int8 | int8 | ~2 GB |
medium.nl | nl | medium | int8 | ~4 GB |
large-v3.nl | nl | large-v3 | int8 | ~14 GB |
small.fr | fr | artkeep/faster-whisper-small-cv11-french | int8 | ~2 GB |
medium.fr | fr | m4th1eu/faster-whisper-medium-french-ct2 | int8 | ~4 GB |
large-v3-french-distil | fr | brandenkmurray/faster-whisper-large-v3-french-distil-dec16 | int8 | ~8 GB |
large-v3.fr | fr | large-v3 | int8 | ~14 GB |
event: started event: log event: segment event: done event: error
Any video or audio format supported by ffmpeg — including MP4, MOV, MKV, WebM, AVI, MPEG, MP3, WAV, M4A, AAC, OGG, FLAC, Opus, WMA.
ffmpeg decodes the file server-side; the browser accepts any video/* or audio/* MIME type.
Optional — currently open (no AUTH_TOKEN set).
JSON status, loaded models, active jobs, memory stats.
Live memory + per-job params, progress, and recent log messages (polled by this page).
Request cancellation of an active job. Queued jobs waiting on the inference lock abort within ~5s; running jobs stop between whisper segments. In-flight ffmpeg may finish first.