Readomi

A reading companion with web, paragraph and subtitle translation.

Readomi is a browser reading companion focused on translation. Translations appear in the page you are reading, in bilingual or translation-only mode, so you can keep reading in place. TRANSLATION BUILT AROUND READING • Translate web pages while preserving their layout and prioritizing visible content. • Switch between bilingual and translation-only views, or restore the original text. • Optionally translate one paragraph by pointing to it and pressing your configured trigger key. • Translate enabled YouTube captions and HTML5 subtitle tracks. Adjust subtitle size, style and position. • Optionally bold the beginnings of English words to suit your reading preference. CHOOSE YOUR OWN SERVICE Use OpenAI, Anthropic, Gemini, DeepSeek, a compatible API or a local model. Configure and test the service manually, or copy the setup instructions for your coding agent to verify a service and prepare its configuration. Discover models and customize prompts and request parameters. MAKE IT YOURS Configure shortcuts and page translation styles. Choose light, dark or system appearance and one of four accent colors. The interface supports several languages, including English and Simplified Chinese. Import and export settings locally. BEFORE YOU START Readomi does not include a hosted translation service. You need your own model service and API key, or a working local model endpoint. Third-party providers may charge for usage and process requests under their own policies. Readomi does not supply credits or guarantee compatibility with every custom endpoint. Page, paragraph and existing subtitle text goes directly to your configured service. After a service is configured, the extension can also send a short page text sample to detect the source language when pages open or change, even before you click Translate. Enabling page context sends the title and main article content to generate a summary; custom prompts can also include page content. Review the privacy policy and your provider's policy. Settings, API keys and caches stay in your browser and are not synced to a developer server. Readomi has no account system, analytics or telemetry. Exported configuration backups contain API keys and custom headers, so keep them private. Hover and video subtitle translation are off by default. Video translation requires existing captions and does not transcribe videos without subtitles. Chrome internal pages and other browser-restricted pages cannot display injected translations. ORIGINS AND CUSTOMIZATION Readomi is adapted from Xuanwo/jiandao and mengxi-ream/read-frog. It primarily removes Read Frog's floating toolbar, text-to-speech, custom AI actions, hosted accounts, configuration sync, analytics and other features, while adding features I need on top of jiandao. I want to keep the features as simple as possible and provide a localized user experience, so I have customized it around my own experience. Another important reason is that I believe, in the AI era, anyone can customize basic capabilities such as translation to suit their needs. If you like the streamlined features Readomi currently offers, you are welcome to use it. If you need features that Readomi intentionally removed from Read Frog, you can use Read Frog or other similar products. Thank you to Read Frog and jiandao for their open-source contributions. Without them, there would be no Readomi. Readomi is independently maintained. It is not affiliated with or officially partnered with either upstream team. Source and support: https://github.com/knothhe/readomi Privacy policy: https://github.com/knothhe/readomi/blob/main/PRIVACY.md