Se-jeremane e mahala
Ho ngola livideo le li-audio tsa Se-jeremane ho ea ho li-text ka AI. Habonolo, ka nepo, le mahala.
Ho sebetsa joang
- E_la ho Motsamaisi oa ho ngola oa Free.ai
- Kopitsa faele ea hau ea audio kapa video ea Se-jeremane
- AI ea rona e fumana Se-jeremane ka ho toba'me e e ngola
- Kopitsa transcript ea hau e le tlhaloso kapa li-subtitles tsa SRT
Se-jeremane Likarolo tsa ho ngola
- ✓E entsoe ke faster-whisper (license ea MIT)
- ✓Ho bona ka ho toba puo ea Se-jeremane
- ✓E tšehetsa MP3, WAV, MP4, M4A, FLAC, le tse ling
- ✓Timestamps le ho romelloa ha lihlooho tse ka tlase (SRT)
- ✓Ha ho na liphelelo tsa boholo ba faele ka li-plans tse lefelloang
- ✓Secha le se sireletsehileng -- lifaele li tlosoa ka mor'a ho sebetsana
Lintlha tsa puo
| Senyesemane | Se-jeremane |
| ISO Code | de |
| Mofuta oa AI | faster-whisper |
| Theko | E le mahala |
Li-languages tse ling
Bona Li-Languages Tseo kaofelaLipotso tse ka etsahalang
Whisper e kholo-v3-turbo e lula ka holimo ho Se-jeremane - ka tlase ho 7% ea palo ea liphoso tsa mantsoe ka litlhahlobo tse tloaelehileng. Ka ho etsa hore ho bolela hore audio ea studio e hloekileng e khutlela morao e le haufi le ho phethahala,'me audio ea puisano e ka sebelisoa le ho hloekisa ho fokolang.(Tier A, under 7% word error rate ka lihlopha tsa benchmark - re phatlalatsa li-tiers tsa WER tse tšepahalang ho feta li-claims tsa thekiso.)
Ha ho joalo - Se-jeremane transcription e nka ho tloha letsatsi le letsatsi token pool mahala pele. Audio theko ka 50 tokens ka metsotsoana, ka hona pool letsatsi le letsatsi Anonymous e koahela lihora tse ling tsa audio ka letsatsi. Signed-in akhaonteng fumana 30,000-token letsatsi le letsatsi pool. Past hore, transcription ke pay-as-you-go, le token top-ups ho tloha $1.
Li-transcripts tsa Se-jeremane li khutlisoa ka UTF-8 e tloaelehileng le ho ngola ka mokhoa o tloaelehileng oa puo.
MP3, WAV, M4A, FLAC, OGG, OPUS, le WEBM li amoheloa ka ho toba. Bakeng sa video (MP4, MOV, MKV) re tlosa li-track tsa audio ka lehlakoreng la mosebeletsi pele re li romella ho Whisper - ha u hloka ho fetola ntho efe kapa efe ka boeona. Pipeline e ts'oanang ntle le puo ea mohloli, ho kenyeletsoa Se-jeremane.
Anonymous uploads cap at roughly 500 MB per file. Signed-in accounts go up to 2 GB. Duration is not a hard limit - long files are chunked automatically (30-second windows with overlap) and stitched back into a single transcript with continuous timestamps. Multi-hour Se-jeremane recordings (podcasts, full lectures, meetings) work fine.
Ho na le - mohlophisi diarization ke ka ho feletseng ka ho feletseng bakeng sa Se-jeremane transcripts tsohle. The output ke segmented joaloka Mohlophisi 1 / Mohlophisi 2 / Mohlophisi 3 le timestamps, kahoo lipotso, paneli lihlooho, le multi-party dibopeho tsa motheo ba tla khutla labelled. Diarization e tsamaea ka model e fapaneng le sebetsa ka tsela e tšoanang ka lipuo tsohle re tšehetsa.
E-na le hoo, re tla kenya URL ho /transcribe/youtube/ bakeng sa YouTube kapa /transcribe/podcast/ bakeng sa podcast feeds (Apple, Spotify, RSS). Re tla kenya audio, re e phethela ka Whisper le language=de, ebe re khutlisa transcript le timestamps le li-labels tsa moqoqi. Lintlha tse tloaelehileng tsa Se-jeremane: podcasts, lingoliloeng, lipotso, le li-long-form YouTube litaba ka Se-jeremane ke li-workloads tse tloaelehileng ka ho fetisisa tseo re li bonang.
Whisper costs about 50 tokens per minute of audio, so a one-hour recording is ~3,000 tokens. Most users never spend anything - the free daily pool of 30,000 tokens covers short clips, voice notes, and one-off podcasts. Beyond it, transcription is pay-as-you-go, with token top-ups from $1.
Ee - li-timestamps tsa boemo ba segment (ka ~10-30 sekontiri ka 'ngoe) le boemo ba mantsoe li fumaneha. Lefatše la mantsoe ke la morao-rao bakeng sa ho romelloa kantle ha li-subtitles tsa VTT/SRT ka hona li-captions li synchronize line-by-line. Ka API beha timestamps="word" ka'mele oa kopo. Li-transcripts tsa Se-jeremane li khutlisoa ka UTF-8 e tloaelehileng le ho ngola ka mokhoa o tloaelehileng oa puo.
Yeah. POST audio (multipart/form-data, field name "file") to /v1/transcribe/ with language=de - or leave out the language parameter to let Whisper auto-detect. Returns JSON with the transcript, segments, timestamps, and speaker labels. Full reference and SDK snippets at /api/.
E-ea — ha transcription e felile, tobetsa ho fetolela kapa ho kopanya tekanyo ho /translate/. Se-jeremane e kopantsoe le lipuo tsohle tse ling tseo re li tšepang (200+). Bakeng sa lihora tsa kopano, transcript e fetisoa ka /summarize/; bakeng sa ho ngola ka letsoho, e romelle ho /voice/tts/ ho etsa molumo ka puo e loketseng.
Whisper e koetlisitsoe ka lihora tse 680K tsa molumo o moholo oa lefats'e, ka hona ho ngolisoa ha Se-jeremane ho matla ho mohala oa morao, li-beds tsa molumo, le ho ngolisoa ha boleng ba mohala. Ho qeta ho thata kapa ho arolelana li-speakers tse ngata ho tla tsoelapele ho senya botsitso.Haeba transcript e sa khone ho sebelisoa, romella lengolo-tsoibila contact@free.ai le faele — re tla lefa tokens'me re sheba hore na e mong oa li-engine tse fapaneng o sebetsana joang le audio ea hau ka ho fetisisa.