Distil-Whisper large-v3
Free.ai (self-hosted)
·
stt
·
~500 tokens per minute
Distil-Whisper large-v3 is a Kutaura-ku-tebhu model built by HuggingFace. Strongest at Real-time transcription, large-volume batch STT.. Self-inotarisirwa pa Free.ai GPUs - inofamba zvakasununguka pazuva rako re token pool (500 tokens per minute). Yakaburitswa pasi peMIT - kushandiswa kwekutengesa kwakabvumirwa pasi peFree.ai.
Usati washandisa
OpenAI-inowirirana REST API. Kugadzira chiratidzo uye kukumbira iyi model mumasekondi.
curl -X POST https://api.free.ai/v1/stt/ \
-H "Authorization: Bearer sk-free-..." \
-H "Content-Type: application/json" \
-d '{"model":"distil-whisper-large-v3","audio_url":"https://..."}'Zvimwe zvinobvunzwa kakawanda
Distil-Whisper large-v3 inoshandura mashoko akataurwa kuita mashoko. Upload a MP3, WAV, M4A, kana video file and Distil-Whisper large-v3 returns the full transcript plus optional SRT/VTT subtitles with timestamps.
Distil-Whisper large-v3 handles dozens of languages - Whisper-family models cover 90+, others vary. Pick "auto-detect" or specify the language for highest accuracy.
Word-error rate i 5-10% pachokwadi English audio, 10-20% panyama kana accented audio. Large variants of the same architecture do meaningfully better on hard cases - pick larger when the audio is rough.
Yeah - chero segment inosanganisira kutanga / kupera timestamps. Export se SRT kana VTT uye nguva map yakananga pane yako video.
Distil-Whisper large-v3 inoita basa redunhu GPUs pamusoro yako zuva nezuva vakasununguka pool yekutanga; $ 5 → 200,000 akabhadharwa tokens mushure meizvozvo. About ~ 500 tokens pamin.
MP3, WAV, M4A, FLAC, OGG, plus video (MP4, MOV, WebM) - isu tinotora audio. Max 500 MB per upload. Longer files? Split with /audio/cut/ or use /v1/stt/batch/.
Munyori diarization ndeimwe yakasiyana kutevedzera - toggle "diarize" pa / transcribe /. Distil-Whisper large-v3 anobata transcription; diarization mazita ese segment ne Munyori 1 / Munyori 2 / etc.
Yes - /batch/ accepts a folder of audio files. Every transcript lands in /account/?tab=history with the original filename. For folder-tree preservation use the API.
Yes — POST your audio to /v1/stt/transcribe/ with model="Distil-Whisper large-v3". Returns JSON with text + segments + word-level timestamps. /api/ has the full reference.
Self-hosted mamodheru kuchengetedza audio paGPU yedu; premium kuuya kuburikidza neDPA. Audio inofanira kugadziriswa mushure mekugovera-window (24h ano, 7d signed-in). Isu hatina kudzidzisa pazvinhu zvako.
Yes - Free.ai grants commercial use of transcripts. You need rights to the audio you uploaded (your own recording, licensed material, or content with consent).
Real-time factor ndeye 0.05-0.2× - 60-minute podcast inoshandura mu3-12 maminitsi. Premium models zvinotora nguva yakareba kuti zviite. Usashandisa bhatani rekumirira kuti uvhure tab.