mengxi-ream/read-frog
🐸 Read Frog - Open-source AI language learning & flashcards, an Immersive Translate alternative | 🐸 陪读蛙 - 开源 AI 语言学习与闪词卡,沉浸式翻译替代
📚 What is Read Frog?
Read Frog is an open‑source browser extension that turns everyday web‑page reading into an AI‑assisted language‑learning experience. It runs inside Chrome, Edge or Firefox and lets you translate, explain, and listen to any text on a page, while also offering tools for building flashcards and custom AI actions.
🎯 Core Capabilities
| Feature | What it does | Why it matters |
|---|---|---|
| Bilingual / Translation‑Only modes | Shows the original text side‑by‑side with its translation, or replaces it entirely. | Lets learners compare directly or focus on a clean reading view. |
| Selection Translation Toolbar | Highlight any text → a floating toolbar appears with Translate, Explain, and Speak buttons. | Instant look‑ups without leaving the page. |
| Context‑Aware Translation | When enabled, the extension sends the page title and a concise Markdown snapshot of the whole page to the AI model. | The AI can disambiguate terms and keep domain‑specific wording accurate. |
| Subtitle Translation | Translates YouTube subtitles in‑place, showing both original and translated captions. | Turns video watching into a language‑learning session. |
| Text‑to‑Speech (TTS) | Uses Microsoft Edge TTS (150+ voices, 80+ languages) to read selected text aloud, with adjustable rate, pitch and volume. | Supports pronunciation practice and listening comprehension. |
| Batch Requests | Groups many translation calls into a single API request, cutting token usage by up to 70 %. | Reduces cost when using paid LLM providers. |
| 20+ AI Providers | Connects to OpenAI, Anthropic, Google Gemini, Groq, Mistral, Ollama, etc., via the Vercel AI SDK, plus free translators like Google Translate and DeepLX. | Gives users flexibility to pick a model that fits their budget or language needs. |
| Custom AI Actions | Users can define prompts, output fields, icons and choose a provider/model, then run the action from the selection toolbar. | Turns the extension into a personal dictionary, summarizer, rewrite tool, or any workflow the learner designs. |
| Flashcards & Spaced Repetition | Save vocab, definitions, examples to Notebase and generate flashcards with customizable templates. The built‑in scheduler shows cards when you’re likely to forget them. | Turns passive reading into active review without extra apps. |
🛠️ How it works under the hood
- The UI is a standard browser‑extension popup/toolbar written in JavaScript/TypeScript.
- Translation and other AI‑driven actions are performed by calling the selected provider’s API (e.g., OpenAI’s
chat/completionsor a free translation endpoint). - For context‑aware translation the extension extracts the page’s DOM, converts it to a concise Markdown representation, and includes that as part of the prompt sent to the model.
- Batch processing aggregates multiple short translation requests into a single larger request, then splits the response back to the original UI elements.
- Text‑to‑speech leverages the public Edge TTS service, handling language detection (simple heuristic or LLM‑based) and chunking long passages at sentence boundaries for smooth playback.
🚀 Getting Started
- Install – Grab the extension from the Chrome Web Store, Edge Add‑ons, or Firefox Add‑ons (links provided in the README).
- Choose a provider – Open the settings, add your API key for a model you prefer, or stick with the free Google/Microsoft translators.
- Start reading – Visit any article, select text, and use the toolbar to translate, hear, or explain it.
- Create flashcards – Click the Notebase button to save vocab for later spaced‑repetition review.
- Customize – Define your own AI actions via the “Custom AI Actions” UI if you need specialized workflows.
🤝 Community & Contribution
- Discord & WeChat – Official channels for support and discussion.
- Contribution guide – The repo includes a
CONTRIBUTING.mdand a detailed guide on how to submit code, documentation, or new AI‑action templates. - Dual licensing – GPL‑v3 for open‑source use, with a commercial license option for closed‑source products (e.g., Meituan Tabbit Browser Team).
📦 Who might benefit?
- Language learners who want immersive, on‑the‑fly translation while browsing.
- Teachers or content creators looking for a quick way to generate bilingual materials.
- Developers interested in building custom AI‑powered browser tools or integrating additional LLM providers.
Read Frog therefore sits squarely in the AI‑enhanced language‑learning space, offering a practical, extensible browser‑based platform that leverages modern LLM APIs, free translation services, and TTS to make reading foreign‑language content an active learning activity.
Related
- Project
- Project
- Project
- Project