AI subtitle generator: automatic subtitles for videos
Drop a video or audio file and AI speech recognition (OpenAI Whisper) writes timed subtitles in about 99 languages. Edit them, then download SRT/VTT or burn them into the video. The AI runs on your own device — your video is never uploaded.
- ✓ No upload
- ✓ No sign-up
- ✓ No watermark
- ✓ Free
Subtitles (click a time to jump there; edit text directly)
How to use
- Drop a video or audio file (MP4, MOV, MKV, MP3, M4A, WAV…).
- Pick the spoken language (or auto-detect) and a model: Fast, Balanced or Accurate.
- Click Generate subtitles. The first run downloads the AI model once (40–250 MB), then it is cached.
- Fix any mistakes in the list, then download SRT / VTT / TXT, or click "Burn into video" to get an MP4 with subtitles.
FAQ
Is it really free and private?
Yes. The Whisper model runs inside your browser using WebAssembly or your graphics card (WebGPU). Your audio is never sent anywhere; only the model files are downloaded from Hugging Face.
How accurate is it?
Accuracy depends on the audio quality and model. "Balanced" works well for clear speech; "Accurate" handles accents, noise and Chinese better but is slower and downloads more. Always proofread names and numbers.
How long does it take?
With a graphics card (Chrome/Edge with WebGPU) a 10-minute video usually takes 1–3 minutes. On CPU only, expect roughly real time with "Fast" and slower with larger models. Keep the tab open while it works.
Can it translate subtitles into English?
Yes. Choose "Translate to English" and Whisper writes English subtitles for speech in any supported language.
What is an SRT file?
SRT is the most common subtitle format. YouTube, Bilibili, Premiere Pro, Final Cut, CapCut, VLC and most players accept it. VTT is the web version of the same thing.
Chinese subtitles come out in Traditional characters?
Whisper sometimes writes Mandarin in Traditional Chinese. Keep "Simplified Chinese output" checked and the subtitles are converted to Simplified characters.