Moonshine Base
Free.ai (self-hosted)
·
stt
·
~500 tokens per minute
Moonshine Base se a Modèl pale-a-tèks ki te bati pa Useful Sensors. Li se pi gwo vil la nan Low-latency live transcription, embedded devices.. Self-hosted sou Free.ai GPUs — kouri gratis kont chak jou ou a token pool (500 tokens pou chak minit). Li libere anba lisans MIT — itilize komèsyal pèmèt sou Free.ai.
Itilize via API
API REST ki konpatib ak OpenAI. Pwodui yon kle epi rele modèl sa a nan kèk segond.
curl -X POST https://api.free.ai/v1/stt/ \
-H "Authorization: Bearer sk-free-..." \
-H "Content-Type: application/json" \
-d '{"model":"moonshine-base","audio_url":"https://..."}'Kesyon ki poze souvan
Moonshine Base transkri son ki pale nan tèks. Upload yon MP3, WAV, M4A, oswa videyo dosye ak Moonshine Base retounen transkript konplè plus optional SRT / VTT sous-titres ak timestamps.
Moonshine Base ka sèvi ak dè dizèn de lang — Whisper-familie modèl kouvri 90+, Parakeet kouvri ~25, lòt varye. Chofe "deteksyon otomatik" oswa endike lang pou pi bon presizyon.
Word-erè pousantaj se 5-10% sou son an Angle pwòp, 10-20% sou son an briyan oswa accented. gwo varyasyon nan menm achitekti fè pi byen sou ka difisil - chwazi pi gwo lè son an se rèd.
Wi - chak segment gen ladann kòmanse/fini timestamps. Ekspòtasyon kòm SRT oswa VTT ak tan mape dwat sou ou vidéo.
Moonshine Base kouri sou GPUs nou menm kont baz done gratis ou chak jou an premye; $5 → 200,000 tokens peye apre sa. Avèk ~500 tokens pa minit.
MP3, WAV, M4A, FLAC, OGG, plus videyo (MP4, MOV, WebM) — nou ekstrè son an. Max 500 MB pou chak upload. Pi long dosye? Divizyon ak /audio/cut/ oswa itilize /v1/stt/batch/.
Diarization pale se yon pase separe — toggle "diarize" sou /transcribe/. Moonshine Base jere transkriptyon an; diarization étiquettes chak pati ak pale 1 / pale 2 / elatriye.
Wi — /batch/ aksepte yon dosye ki gen dosye son. Chak transkripte fini nan /account/?tab=history ak non dosye orijinal la. Pou konsève arbr-dosye a, itilize API a.
Wi — POST ou son an nan /v1/stt/transcribe/ ak model="Moonshine Base". Retounen JSON ak tèks + segments + word-level timestamps. /api/ gen referans konplè.
Modèl ki òganize tèt yo kenbe son sou GPU nou yo; premium pase atravè ak yon DPA. Son an se efase apre fenèt pataje a (24h anon, 7d ouvri sesyon). Nou pa antrene sou enpòte ou yo.
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).
Faktori tan reyèl la se apeprè 0.05–0.2× — yon podcast 60-minute transkri nan 3–12 minit. Modèles Premium yo souvan fini pi vit.