AI · Free 30 min/day · Pro for long-form
Transcribe Video to Text
Drop a video, get a clean transcript. Audio is extracted privately in your browser, then transcribed on our servers using Whisper-large-v3-turbo.
Checking browser capabilities...
Drop your video or audio here
MP4, MOV, MKV, WebM, MP3, WAV, M4A — up to 2GB. Audio extracted in your browser, transcribed on our servers.
Why this tool
Audio stays minimal
We never upload your video — only a 16 kHz mono MP3 distilled from it. A 1-hour clip becomes ~28 MB, even on slow uplinks.
State-of-the-art accuracy
Backed by Whisper-large-v3-turbo on Groq, with multilingual support and timestamps for every segment.
Free tier that's actually useful
Up to 30 minutes per day on the free tier; Pro and Business unlock long-form transcripts and priority queue.
How to transcribe a video
- 1Drop your video (MP4 / MOV / MKV / WebM / MP3 / WAV / M4A) into the uploader.
- 2We extract the audio in your browser at 16 kHz mono — fast and small.
- 3The audio is sent to our server and transcribed by Whisper-large-v3-turbo.
- 4Read, copy, or export the transcript. Editor and SRT/VTT/MD/JSON exports are coming soon.
Frequently asked questions
- Does my full video get uploaded?
- No. Only the audio track is extracted in your browser, downsampled to 16 kHz mono MP3, and uploaded — typically less than 1 MB per minute. We never see your raw video.
- Which languages are supported?
- Whisper-large-v3-turbo supports 90+ languages including English, Spanish, Mandarin, Japanese, French, German, Portuguese, Russian and Arabic. Auto-detection is on by default.
- How accurate is it?
- Whisper-large-v3-turbo is on par with the largest open transcription models — typical word-error rate on clean English audio is below 5%. Noisy or heavily accented audio is harder.
- What's the maximum length?
- Free tier caps at 30 minutes per file (30 minutes / day total). Pro lifts to 10 hours / month with up to 3 hours per file. Business gets 30 hours / month and a priority queue.
- Can I get timestamps?
- Yes. The result includes per-segment start / end timestamps. The editable transcript view and SRT / VTT / Markdown / JSON exports are arriving in the next release.
- Do you store my transcripts?
- Free tier transcripts are kept for 24 hours so you can re-download them; Pro 7 days; Business 30 days. You can delete any transcript from the dashboard at any time.
- Will the audio I upload be used to train models?
- No. Audio is sent to Groq for inference only and is not retained beyond the request lifetime; Clapr does not retain the audio after transcription completes.