Moonshine Base
Free.ai (self-hosted)
·
stt
·
~500 Token minute
Moonshine Base bụ a ngwe-na-akwụkwọ móòdù e mepụtara site na Useful Sensors. Ọrụ na Low-latency live transcription, embedded devices.. Self-hosted na Free.ai GPUs - na-agba ọsọ n'efu megide ụbọchị gị token pool (500 tokens nkeji). E wepụtara ya n'okpuru MIT - iji ahịa ekwenyela na Free.ai.
Jiri site na API
OpenAI-na-akpaghị aka REST API. Kewapụta kii na-akpọkwa model a n'ime sekọnd.
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://..."}'Ajụjụ ndị a jụrụkarị
Moonshine Base na-ewepụ ụda e kwuru n'ime ngwe. Bipụta MP3, WAV, M4A, mọọbụ faịlụ vidio na Moonshine Base na-ewepụ ntụgharị zuru ezu nakwa SRT/VTT subtitles na-ahọrọ na oge.
Moonshine Base handles dozens of languages - Whisper-family models cover 90+, others vary. Pick "auto-detect" or specify the language for highest accuracy.
Nhazi okwu-isiokwu bụ 5-10% na ụda Bekee dị ọcha, 10-20% na ụda na-anụ ọkụ ma ọ bụ ụda akọwapụtara. Ụdị dị ukwuu nke ụda ahụ na-eme ka ọ dị mma n'ihe ndị dị njọ - họrọ nnukwu mgbe ụda ahụ bụ n'okporo ụzọ.
Ee - segmenti ọbụla na-agụnye oge ngwụcha/ọgwụcha. Ekpughe ya dịka SRT mọọbụ VTT na oge map n'ụzọ ziri ezi na vidio gị.
Moonshine Base na-agba ọsọ na GPUs anyị onwe anyị megide ụbọchị gị n'efu pool mbụ; $ 5 → 200,000 akwụ ụgwọ tokens mgbe ahụ. N'ihe banyere ~ 500 tokens kwa nkeji.
MP3, WAV, M4A, FLAC, OGG, nakwa vidiyo (MP4, MOV, WebM) - anyị na-ewepụ ụda. Max 500 MB kwa ibudata. Faịlụ ogologo? Kpọchie na /audio/cut/ ma ọ bụ jiri /v1/stt/batch/.
Onyeọsụsụ diarization bụ n'otu n'otu n'ime - toggle "diarize" na /transcribe /. Moonshine Base na-elekọta ntụgharị; diarization na-egosipụta segmenti ọbụla na Onyeọsụsụ 1 / Onyeọsụsụ 2 / wdg.
Ee - /batch/ na-anabata nsomebefaịlụ nke faịlụ ụda. Transcript ọbụla na-abanye na /account/?tab=history na aha faịlụ mbụ. Maka nchekwa nsomebefaịlụ-tree jiri API.
Ee - POST ụda gị na /v1/stt/transcribe/ na model="Moonshine Base". Na-eziga JSON na ngwe + segments + oge-iheomume nke okwu-ọsọ. /api/ nwere nlekọta zuru ezu.
Self-hosted models na-echekwa ụda na GPUs anyị; premium na-aga site na DPA. Ọdịdị a ehichapụla mgbe windo n'ozuzu ya (24h anon, 7d banye-ime). Anyị anaghị arụ ọrụ na init gị.
Ee - Free.ai na-enye ikike iji n'ọrụ n'ihe n'ime nke transcripts. Ichọrọ ikike n'ihe n'ime nke ọwa ịgbanyela (n'ihe n'ime gị, ihenhọrọ na-enye ikike, mọọbụ ihenhọrọ na-enye ikike).
Real-time factor bụ ihe dịka 0.05-0.2× - 60-minute podcast transcribes na 3-12 minit. Premium models often finish faster. Use the queue button to close the tab.