Distil-Whisper large-v3
Free.ai (self-hosted)
·
stt
·
~500 token kasta minute
Distil-Whisper large-v3 waa a qaab hadal-u-qoraal dhisay HuggingFace. Ugu xoog badan ee Real-time transcription, large-volume batch STT.. Is-hoosaysiinta Free.ai GPUs - bilaash ayaa ka socda ishaada maalinlaha ah (500 tokens daqiiqo kasta). Soo baxay hoos MIT — isticmaalka ganacsi ee la oggol yahay on Free.ai.
isticmaalka API
OpenAI-ku habboon REST API. abuuro fure iyo wicitaan noocan ah daqiiqado.
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://..."}'Su'aalaha badanaa la isweydiiyo
Distil-Whisper large-v3 ku soo celin audio hadalka in qoraalka. Upload MP3, WAV, M4A, ama faylka video iyo Distil-Whisper large-v3 ku soo celin kartaa qoraalka buuxa iyo sidoo kale ikhtiyaari ah SRT / VTT subtitles la timestamps.
Distil-Whisper large-v3 xakamayn kartaa boqolaal afaf — Whisper-qoyska qaababka daboolaan 90+, Parakeet daboolaa ~25, kuwa kale oo kala duwan. Pick "auto-ka ogaado" ama si gaar ah u qeexida afka ugu sarreeya ee saxnaanta.
Qiimaha eray-wax-ka-beddelka ah 5-10% on audio English nadiif ah, 10-20% on audio xawaare ama accented. noocyada waaweyn ee dhismaha isku mid ah si macno leh u fiican u sameeyaan kiisas adag - dooro weyn marka audio waa adag.
Haa - qayb kasta oo ka mid ah bilow / dhamaadka timestamps. dhoofinta sida SRT ama VTT iyo mar kasta oo madal si toos ah u ku saabsan video.
Distil-Whisper large-v3 ku socda GPUs our gaarka ah ka dhanka ah maalin kasta free pool hore; $ 5 → 200,000 ka dibna lacagta la siiyay.
MP3, WAV, M4A, FLAC, OGG, iyo sidoo kale video (MP4, MOV, WebM) — waxaan soo saarno audio. Max 500 MB per upload. Faylal dheer? kala qaybsan /audio/cut/ ama isticmaal /v1/stt/batch/.
Speaker diarization waa ka qayb qaadanaya kala duwan — toggle "diarize" on / transcribe /. Distil-Whisper large-v3 xakamaynaysaa soo gudbinta; diarization labels qayb kasta oo la Speaker 1 / Speaker 2 / iwm
Haa — / batch / aqbalaa folder files audio. Dhammaan qoraalka dhulka / account /? tab = taariikhda leh magac file asalka ah. U dir folder-geed ilaalinta isticmaalka API.
Haa — POST aad audio in /v1/stt/transcribe/ la model="Distil-Whisper large-v3". JSON la qoraalka + qaybaha + eray-tayada timestamps ku soo laabtaa. /api/ waxaa ku qoran soo jeedinta buuxda.
Self-hosted models haystaa audio on our GPUs; premium gudbaan iyada oo DPA. Audio waa la tirtiri doonaa ka dib markii share-dhismaha (24h anon, 7d signed-in).
Haa — Free.ai waxay siisaa isticmaalka ganacsi ee qoraalada. Waxaad u baahan tahay xuquuqda maqalka aad soo dejisay (rekorkaaga, waxyaabaha la siiyay liisan, ama waxyaabaha la oggolaaday).
Waqtiga dhabta ah waa in ka badan 0.05-0.2 × - 60-daqiiqo podcast ku qorto 3-12 daqiiqo. Premium models badanaa dhamaystiraan si deg deg ah.