OpenSubs

OpenSubs is a browser-based AI subtitle generator that transcribes, translates, styles and burns subtitles into video without ever uploading the file to a server.

What is OpenSubs?

OpenSubs is a free, open-source subtitle generator that runs entirely inside your web browser. You drop in a video file, and it produces a finished MP4 with burned-in subtitles, or exports .srt, .vtt and .ass files. Speech recognition uses Whisper running on your own GPU, and the video is read through an object URL so it never leaves your machine. The project is published under AGPL-3.0 by the OpenSubs team.

Key Features

• On-device Whisper transcription — pick a model between 40 MB (Tiny, English), 80 MB (Base) and 250 MB (Small); it downloads once and is cached.
• Automatic language detection — the spoken language is re-read from the audio every four seconds, so videos that switch between two languages are transcribed in both and merged into one subtitle file.
• Translation into 20 languages — Chrome's built-in on-device translation is free; bring your own Claude, OpenAI or DeepL key, or pay in credits for OpenSubs' backend model.
• Twelve subtitle style presets — from clean broadcast captions to bold social styles, previewed over your own footage, with word-by-word grow-and-glow highlighting.
• Burn subtitles into MP4 — encoding happens in the browser with WebCodecs; alternatively export .srt, .vtt or .ass.
• Trimming and resolution control — adjust the clip and output resolution before export.
• Three companion builds — a browser extension (Chrome, Edge, Firefox), a desktop app (macOS, Windows, Linux) and a command-line tool, all sharing the same Rust engine.
• No account, no upload, no watermark — every step runs locally unless you explicitly choose a cloud option.

Who is it for?

• Content creators who need captions burned into short-form video without sending footage to a third-party service.
• Privacy-conscious users handling sensitive or unreleased footage who cannot upload it to a cloud transcriber.
• Developers and scripters who want the same engine via CLI for batch jobs or CI pipelines.
• Multilingual creators working with videos that mix two spoken languages in one recording.

What can you do with OpenSubs?

• Subtitle a video in the browser: drop a file, choose a Whisper model and language, generate subtitles, style them and export a burned-in MP4 or an .srt/.vtt/.ass file.
• Translate existing subtitles: run on-device Chrome translation for free, or supply your own Claude, OpenAI or DeepL API key for long dialogue.
• Subtitle a video already playing in a tab: the browser extension generates subtitles over whatever is playing, as it plays.
• Batch-process long files: the desktop app drives ffmpeg (built with the ass filter) for long recordings and batches; the CLI does one command per file.

How does OpenSubs work?

The workflow has four steps, none of them on a server: transcribe the speech with Whisper in the browser, translate the text (on-device by default), style it with one of twelve presets, then burn it into the video with WebCodecs or export a subtitle file. The video is read through an object URL and decoded locally; only subtitle text can leave your machine, and only if you pick a cloud translator.

Pricing

OpenSubs is free for everything that runs on your machine — speech recognition, on-device translation, all twelve styles, word-by-word highlighting, trimming, burning and SRT/VTT/ASS export. Cloud translation on OpenSubs' backend costs $5 for 1,000 credits, with the price of a job shown before you run it, failed jobs costing nothing, and credits that never expire. Your own API keys are also supported at no charge from OpenSubs.

FAQ

Is OpenSubs free?

Yes. Transcription, on-device translation, all twelve subtitle styles, word-by-word highlighting, trimming, burning subtitles into MP4 and SRT/VTT/ASS export are fr