Udemy TTS
A browser extension that translates Udemy lecture subtitles with an LLM and reads them aloud in real time, in sync with the video.
Features
- Multiple translation providers Google Gemini, OpenAI, OpenRouter, or your own OpenAI-compatible endpoint, plus free on-device machine translation in Chrome and Edge. Only low-cost, fast models are offered as presets.
- Translator personality Standard, concise, and friendly presets, or write your own custom instructions for how translations should read.
- Accurate translation Indexed JSON with structured output keeps every line matched to its source. Gaps are retried as whole chunks, and playback never stops even if a translation is missing.
- Translation cache Results are cached per lecture, so revisiting a course doesn't re-translate it. Review, delete, or clear the cache from the options page.
- Two speech engines Edge TTS by default, or VOICEVOX for a different Japanese voice, running on your own machine.
- Automatic speed adjustment Playback speed adapts continuously to keep the narration in sync with the video, and is guaranteed to catch up before the lecture ends.
- Pause & resume that just works Pausing the video pauses the narration at the same spot. A watchdog recovers automatically if a media key accidentally stops only the audio.
- Pronunciation overrides A personal dictionary plus a lecture-specific list the LLM generates automatically, so technical terms and names are read correctly.
- Subtitle overlay Optionally show the translated (or original) text over the video, independent of narration, and drag it wherever you like.
- Built-in setup guide Step-by-step instructions for getting an API key and configuring billing for each provider, bundled right into the extension.
How it works
- Open the extension's options page, choose a translation provider, add your API key, and run "Test connection".
- Open any Udemy lecture — a control button is added to the video player.
- Use the panel to toggle translation, narration, and the subtitle overlay, and to adjust volume and speed.
Works on both Chrome and Firefox (Manifest V3). Source code and issue tracker are on GitHub.