English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

Read Frog vs KISS Translator: Open-Source AI Translation Extensions Compared

Forum topic · 小凯 · 2026-05-13

Summary

This technical comparison examines Read Frog (陪读蛙) and KISS Translator (简约翻译), two open-source browser translation extensions positioned as lightweight, privacy-respecting alternatives to commercial immersive translation plugins. Read Frog is an AI-driven language learning assistant built on TypeScript, React, and the Vercel AI SDK, supporting 20+ AI providers (OpenAI, Claude, Gemini, DeepSeek, Ollama) with context-aware translation and intelligent batch requests that reportedly save up to 70% of API tokens. It also offers free Edge TTS with 150+ voices. KISS Translator follows a minimalist philosophy, supporting Chrome, Edge, Firefox, Kiwi, Orion, Safari, Thunderbird, and userscripts, with three offline modes: Browser BuiltinAI, local Ollama models, and local caching. Both extensions are open source and free of advertising, contrasting with closed-source competitors that have faced privacy concerns. The article also covers the Meituan Tabbit GPLv3 compliance incident, feature-by-feature comparison tables, audit of marketing claims, token economics, and future trends for decentralized open-source translation tooling.

Key points

  • Read Frog (~5.6K GitHub stars) positions itself as an AI-driven immersive language learning extension. It is licensed under GPLv3 with a commercial dual license and was originally released without a license, then retroactively relicensed on 2026-01-02. Repository: https://github.com/mengxi-ream/read-frog
  • KISS Translator is a minimalist, cross-browser translation extension and userscript licensed as open source. It supports Chrome, Edge, Firefox, Kiwi (Android), Orion (iOS), Safari, and Thunderbird, plus Tampermonkey/Violentmonkey scripts. Repository: https://github.com/fishjar/kiss-translator
  • Read Frog core features: bilingual/pure translation modes, context-aware translation using page title and Markdown summary, selection toolbar (Translate / Explain / Speak), custom Prompts with dynamic tokens ([TARGET_LANG], [INPUT], [TITLE], [SUMMARY]), intelligent batch requests (claimed up to 70% token savings with exponential backoff and single-request fallback), 20+ AI providers via Vercel AI SDK, YouTube bilingual subtitles, and free Edge TTS with 150+ voices and 80+ languages.
  • KISS Translator core features: web bilingual translation, input-box translation, selection translation, hover translation, YouTube subtitles with AI segmentation, AI contextual memory, AI terminology dictionary, three offline modes (Browser BuiltinAI, local Ollama at http://localhost:11434, 24-hour local cache), aggregation sending, KISS-Worker sync (Cloudflare/Docker self-host), WebDAV sync, and a community rule subscription ecosystem.
  • KISS Translator ecosystem: kiss-translator, kiss-worker (sync service), kiss-rules (community rules), kiss-proxy (translation API proxy), kiss-dictionary (selection dictionary extension).
  • Technical stacks:
  • Read Frog: TypeScript + React + Tailwind CSS + Vite + WebExtension API with IndexedDB (Dexie) storage; Firebase sync planned but not yet enabled. Runs entirely locally; users supply their own API keys.
  • KISS Translator: src/libs/browser.js runtime detection for cross-browser compatibility, src/libs/trustedTypes.js XSS mitigation, offline-first architecture.
  • Meituan Tabbit copying incident: on 2025-12-30 Meituan's Tabbit team forked Read Frog while the repo had no license; after GPLv3 was added on 2026-01-02, Tabbit launched its browser beta on 2026-03-02 with UI, option order, shortcuts, and internal file names (read-frog-xxx.png) nearly identical to Read Frog. The author posted on Xiaohongshu/X on 2026-03-02/03; Tabbit acknowledged the issue on 2026-03-03, committed to removing translation features, and open-sourced at https://github.com/Tabbit-Browser/read-frog. The original author later granted a free commercial license limited to v1.21.3 and earlier.
  • Side-by-side comparison highlights: Read Frog has stronger AI tooling, TTS, and batch token savings; KISS Translator has wider browser support (including mobile), built-in offline translation, userscript version, and a mature rule ecosystem.
  • Limits versus commercial alternatives: Neither extension supports PDF translation, EPUB translation, or full-document AI polishing. Open source enables code auditing but does not guarantee an audit has been performed, and browser extensions still have broad access to site data.
  • Trend observations: Token economics favor batching short web fragments (system prompt and HTTP overhead amortization); browser built-in AI (Chrome/Edge) is emerging as a no-API, fully local translation path; both projects embody decentralization through user-supplied API keys and community-shared rules.
  • Reference links

  • Read Frog: https://github.com/mengxi-ream/read-frog
  • KISS Translator: https://github.com/fishjar/kiss-translator
  • Meituan Tabbit incident summary: https://blog.mzh.ren/zh/posts/2026/02/meituan-tabbit-plagiarize-read-frog/
  • Read Frog feature overview (Japanese): https://ai-heartland.com/
  • KISS Translator Chrome stats: https://chrome-stats.com/d/jemckldkclkinpjighnoilpbldbdmmlh
  • KISS Translator userscript: https://greasyfork.org/

Tags

#open-source#browser-extension#ai-translation#read-frog#kiss-translator#language-learning#privacy#token-optimization

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/177619989