Free Cantonese Transcription
Kushandura Cantonese audio uye video kuita tebhu neAI. Yakanaka, yakakwana uye yemahara.
Maitiro Ekushanda
- Tora Free.ai Transcriber
- Kuisa yako Cantonese audio kana video faira
- Our AI otomatiki anowana Cantonese uye anonyora
- Dhawunirodha transcript yako setext kana SRT subtitles
Cantonese Transcription Features
- ✓Yakagadzirwa ne faster-whisper (MIT licensed)
- ✓Automatic Cantonese rurimi detection
- ✓Supports MP3, WAV, MP4, M4A, FLAC, uye zvimwe
- ✓Time stamps uye subtitle export (SRT)
- ✓Hapana faira resize limits pazvirongwa zvakabhadharwa
- ✓Private and secure -- files are deleted after processing
Mamwe
| Chirungu | Cantonese |
| ISO Code | yue |
| AI Model | faster-whisper |
| Mutengo | _Dambudziko: |
Mamwe Matauro
View All LanguagesFAQ
Cantonese ndeimwe yemitauro ine zvishoma zvinodiwa zve Whisper - large-v3-turbo inomira pamusoro pe25% yemashoko ematambudziko, dzimwe nguva zvakakwana pamusoro. Iyo transcript inobatsira kune kutsvaka uye gist asi haifanire kutariswa seyekuburitswa kwazvino. Kana imwe yepamusoro-yekunyatsoita injini ichiwanikwa ye Cantonese isu tinoibatanidza otomatiki.(Tier D, over 25% word error rate pa benchmark sets — isu tinoburitsa zvakarurama WER tiers kupfuura zvekutengesa zvikumbiro.)
Yeah — Cantonese transcription inotora kubva kune yako yezuva nezuva yemahara token pool yekutanga. Audio inodhura nezve 50 tokens paminetsi, saka iyo anonyyme yezuva nezuva pool inosvika pamazuva mashoma eaudio pazuva. Akaunti akanyorwa anowana 30,000-token yezuva nezuva pool. Pashure pezvo, transcription ibhadhara-se-unoshandisa, ne token top-ups kubva pa $ 1.
Cantonese transcripts are returned in standard UTF-8 with the language's normal spelling.
MP3, WAV, M4A, FLAC, OGG, OPUS, uye WEBM anobvumirwa zvakananga. Kuti video (MP4, MOV, MKV) isu kubvisa audio track server-side pamberi kutumira kuti Whisper - iwe hauna kufanira kushandura chero chinhu iwe pachako. Same pipeline pasina kuenderana mutauro wemutauro, kusanganisira Cantonese.
Anonymus 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 Cantonese recordings (podcasts, full lectures, meetings) work fine.
Yeah - munyori diarization iri pa default yese Cantonese transcript. The output is segmented as Speaker 1 / Speaker 2 / Speaker 3 with timestamps, so interviews, panel discussions, and multi-party meetings come back labeled. Diarization runs on a separate model and works the same across all languages we support.
Yes - paste the URL into /transcribe/youtube/ for YouTube or /transcribe/podcast/ for podcast feeds (Apple, Spotify, RSS). We download the audio, run it through Whisper with language=yue, and return the transcript with timestamps and speaker labels. Typical Cantonese content: lectures, interviews, voice notes, and YouTube content in Cantonese all work — pedza URL mu /transcribe/youtube/ kana wedzera faira zvakananga.
Whisper inodhura ma tokens 50 paminuutti, saka kurekodha kweawa imwe chete kunodhura ma tokens 3,000. Vazhinji vanoshandisa havabhadhare chero chinhu — iyo yemahara yezuva nezuva pool ye30,000 tokens inobhadhara mavhidhiyo, mavoice notes, uye podcasts. Kunyangwe, transcription inoita sekuti unobhadhara, nechikwereti chekubhadhara kubva pa$1.
Yeah — both segment-level (every ~10-30 seconds) and word-level timestamps are available. Word-level is the default for VTT/SRT subtitle export so the captions sync line-by-line. On the API set timestamps="word" in the request body. Cantonese transcripts are returned in standard UTF-8 with the language's normal spelling.
Yeah. POST audio (multipart/form-data, field name "file") to /v1/transcribe/ with language=yue — or omit 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/.
Yeah — kana transcription yapera, tinya Translate kana kuti nyora mu /translate/. Cantonese inowirirana nesemwe rurimi tichitsigira (200+). Kuti uite misangano, nyora mu /summarize/; kuti uite dub, nyora mu /voice/tts/ kuti uite mashoko mu rurimi rwaunoda.
Whisper's noise training inobatsira zvakanyanya paiyi tier — iyo bottleneck ndeyekuwanda kwe Cantonese audio Whisper yaona panguva yekudzidzisa, kwete noise. Clean studio audio still beats noise audio, but neither will reach the accuracy you'd get on a high-resource language.Kana transcript ichidzokera pasina kushandiswa, email contact@free.ai nefile - isu tichadzosera ma tokens uye tichatarisa kana imwe injini imwechete inokwanisa kudzora audio yako zvakanaka.