Free AI Subtitle/Captions Generator & Editor: SRT, VTT, Burn
Upload Video or Audio
*Auto-generate subtitles for MP4, MOV, MKV, WebM, MP3, WAV & more
Choose AI Model
The model downloads once and is cached for next time.
Caption Status
How to Generate Captions for your Video?
- Add your files - Click "Choose Files" or drag and drop one or more videos (or audio files). Batch captioning is supported
- Pick an AI model - Fast for quick drafts, Balanced for most videos, Accurate for difficult audio. Optionally set the spoken language or enable translation to English
- Generate - Click "Generate Captions"; the Whisper AI model transcribes with word-level timestamps, entirely on your device
- Edit & export - Fix any words in the built-in editor, then download captions as SRT, VTT or plain text
- Optional: burn-in - Permanently render the subtitles onto the video ("hardcode") and download the captioned MP4
Why Use Our AI Caption Generator?
- 100% Free - No hidden costs, no signup, no watermarks, no per-minute limits
- Private by Design - Speech recognition runs entirely in your browser using OpenAI's Whisper model (via WebAssembly / WebGPU). Your video and audio are never uploaded, transmitted, or stored on any server
- Real Word-Level Timestamps - Captions are built from per-word timings, so cues line up naturally with speech instead of drifting
- Batch Transcription - Queue multiple videos and get an SRT/VTT for each in one go
- Subtitle Burn-In - Hardcode the captions onto the video right in your browser using hardware-accelerated encoding, and download a ready-to-share MP4
- 99 Languages - Whisper understands ~99 languages, and can also translate speech into English captions
Frequently Asked Questions
How do I generate subtitles for a video online for free?
Add your video, click Generate Captions, and download the result as SRT or VTT. This free subtitle generator runs OpenAI's Whisper speech-to-text model directly in your browser, so you can auto-generate captions with accurate timestamps without installing software or creating an account.
Is my video uploaded to a server?
No. The AI transcription, caption editing, and subtitle burn-in all happen locally on your device using WebAssembly and WebGPU. The only downloads are the AI model files themselves (fetched once from a public CDN and cached) — your media never leaves your computer.
How big are the AI models that get downloaded?
Each preset downloads its own copy of the Whisper model the first time you use it, and each model is stored separately. On browsers with GPU acceleration (WebGPU in Chrome and Edge): Fast (Whisper Tiny) is about 120 MB, Balanced (Whisper Base) about 200 MB, and Accurate (Whisper Small) about 400 MB. On browsers without WebGPU, smaller compressed versions are used instead: roughly 45 MB, 80 MB and 245 MB. A model downloads only once — it is saved in your browser's local cache and reused on every future visit until you delete it. Once a model has been downloaded, a delete button appears in the tool so you can free up the space whenever you like.
How accurate are the AI-generated captions?
Whisper is one of the most accurate open speech-recognition models available. Clear speech typically transcribes with very few errors on the Balanced model; noisy audio, heavy accents, or specialised vocabulary benefit from the Accurate model. You can fix any word in the built-in caption editor before exporting.
What is the difference between SRT and VTT?
SRT (SubRip) is the most widely supported subtitle format — video editors, players, YouTube, and social platforms all accept it. VTT (WebVTT) is the web-native format used by HTML5 video players and most online video hosts. This tool exports both, plus a plain-text transcript.
Can I hardcode (burn) the subtitles into the video?
Yes. The burn-in option renders the captions permanently onto the video frames using your browser's built-in hardware-accelerated video encoder (WebCodecs), then gives you an MP4 download. On most machines this runs faster than real time, and the original audio track is passed through untouched whenever possible.
Which languages are supported?
Whisper supports about 99 languages including English, Hindi, Spanish, French, German, Portuguese, Arabic, Chinese, Japanese and Korean. Leave the language on Auto-detect, or set it explicitly for better results. You can also translate any supported language into English captions.
How long can my video be? Can I caption a 1GB video?
Transcription has no fixed length limit — long files are processed in 30-second chunks, so even hour-long recordings work if your device has enough memory. For subtitle burn-in, the captioned video is assembled in your device's memory, so files up to around 2 GB work best on a typical computer.
Which browser do I need?
A recent version of Chrome, Edge, or another Chromium-based browser works best (and can use your GPU for faster transcription). Firefox and Safari 16.4+ also work using the WebAssembly engine. Subtitle burn-in additionally needs WebCodecs video encoding, available in Chrome, Edge, Safari 16.4+ and recent Firefox. The tool will tell you if something isn't supported.
Is this tool privacy compliant?
This tool is built to be private by design: it performs all processing locally in your browser and does not transmit, collect, or store your video, audio, or transcripts anywhere. Because no media ever reaches a server, your content is never exposed to us or any third party.
Is there a more powerful option than this free client-side tool?
Yes. This page is a free tool that runs entirely on your device, which keeps it private but limits it to your computer's speed and memory. VdoCipher's video hosting platform includes a far more capable server-side subtitle generation service: it transcribes long videos and large libraries quickly on dedicated infrastructure, with higher accuracy and no load on your device. And because the AI models are self-hosted on VdoCipher's own servers, your videos and transcripts are never sent to any third-party AI provider — your data stays within VdoCipher's secure infrastructure.
Terms & Privacy
- This free caption generator is provided for general use on an “as is” basis. AI transcription can contain mistakes; we do not guarantee accuracy, browser compatibility, or fitness for any purpose. Please review captions before publishing and keep a copy of your original file.
- All processing happens locally in your browser. Your video, audio, and transcripts are never uploaded to, stored on, or transmitted to any server. We have no access to the content you caption.
- By using this tool, you confirm that you have the full legal right and permission to use, transcribe, and modify the file. You are solely responsible for the content you process and for ensuring your use complies with all applicable laws.
- Because nothing is uploaded, no media data is collected. Basic anonymous technical data (such as browser type and error logs) may be used for analytics and improving the tool.