AI Translator

Description

Free, secure AI translation with no middleman. Web pages, academic papers (math/LaTeX preserved), and real-time YouTube captions. ✨ AI Translator — Your Key, Your Data, Your Rules ✨ AI Translator — 你的 Key，你的数据，你的规则 One extension, the whole reading workflow: web pages, papers, videos, images, PDFs and comics. 一个扩展，覆盖完整的阅读工作流：网页、论文、视频、图片、PDF 和漫画。 💰 1. Free to Start, Free to Scale — BYOK 💰 1. 零门槛起步，成本自己掌控 — BYOK 模式 Works out of the box with Chrome's built-in translator — completely free, no API key, no sign-up. When you want higher quality, bring your own key: 开箱即用 Chrome 内置翻译引擎——完全免费，无需 API Key，无需注册。想要更高质量时，接入你自己的 Key： ✅ BYOK (Bring Your Own Key): pay only for actual usage, no middleman markup ✅ BYOK（Bring Your Own Key）：按实际用量付费，没有中间商赚差价 ✅ Supports OpenAI, Claude, Gemini, DeepSeek, OpenRouter and any OpenAI-compatible API ✅ 支持 OpenAI、Claude、Gemini、DeepSeek、OpenRouter 及任何 OpenAI 兼容 API ✅ Auto-adapts request parameters per model generation (GPT-5, Gemini 3, Claude...) — configure once, it just works ✅ 自动适配各代模型的请求参数（GPT-5、Gemini 3、Claude 等）——配置一次，直接可用 ✅ Ollama / LM Studio local deployment supported: run everything at zero cost 🆓 ✅ 支持 Ollama / LM Studio 本地部署，全程零成本运行 🆓 🌐 2. Translate Everything You Read 🌐 2. 你读到的一切，都能翻译 📄 Full-page translation: side-by-side display below the original, page layout untouched 📄 整页翻译：译文原位对照显示在原文下方，不打乱网页排版 ✏️ Selection & hover translation: select or hover to translate just what you need ✏️ 划词与悬停翻译：选中即译、悬停即译，只翻你需要的部分 ⌨️ Input translation: write in your language, send in theirs — reply to emails and forums directly ⌨️ 输入框翻译：用母语输入，一键转成目标语言，直接回复外文邮件和论坛 🎬 Video subtitle translation: bilingual subtitle overlay on YouTube and any site with standard subtitle tracks — draggable, resizable, and we never touch subtitles you didn't turn on 🎬 视频字幕翻译：YouTube 及任何使用标准字幕轨道的网站，双语字幕悬浮显示——可拖动、可缩放，且绝不擅自开启你没有打开的字幕 🖼️ Image OCR: right-click any image (or draw a region) to extract its text, then translate with one click. On-device engine runs offline with no API key; a vision model handles photos and stylized type 🖼️ 图片文字识别：右键任意图片（或框选区域）提取文字，一键翻译。本地引擎离线运行、无需 API Key；照片和艺术字可切换视觉模型 🔊 Read aloud: TTS for both source and translation 🔊 朗读功能：原文、译文均可语音朗读 🔒 3. Privacy by Architecture 🔒 3. 隐私来自架构，而非承诺 For all web translation, your data goes directly from your browser to the AI endpoint you chose — there is no server of ours in between: 所有网页翻译中，数据从你的浏览器直连你自己选择的 AI 端点——中间没有我们的服务器： 📱 Your Browser ──→ 🤖 Your API ──→ 📄 Your Translation 📱 你的浏览器 ──→ 🤖 你的 API ──→ 📄 你的翻译结果 🚫 We don't collect, store, or analyze your content 🚫 不收集、不存储、不分析你的任何内容 🏠 With local deployment (Ollama / LM Studio), data never leaves your machine or intranet 🏠 本地部署（Ollama / LM Studio）模式下，数据完全不出本机/内网 🖼️ Local image OCR runs entirely on your device — works offline 🖼️ 本地图片识别完全在你的设备上运行——离线也能用 ℹ️ Transparency: only PDF and comic translation are processed on our servers (they require re-typesetting), and both are opt-in, sign-in features ℹ️ 透明说明：仅 PDF 翻译与漫画翻译在我方服务器处理（因需要重新排版），且均为需登录的可选功能 📐 4. Built for Academic Papers 📐 4. 为学术论文而生 This is our biggest differentiator ⚡: 这是我们最大的差异化优势 ⚡： 🧮 Math formulas: traditional tools break them ❌ — we auto-detect MathJax / KaTeX, protect with placeholders, and precisely restore ✅ 🧮 数学公式：传统工具直接翻译破坏 ❌ —— 我们自动识别 MathJax / KaTeX，占位符保护，翻译后精准还原 ✅ 📝 LaTeX symbols: garbled or lost elsewhere ❌ — $...$, $$...$$, \begin{equation} structures fully preserved here ✅ 📝 LaTeX 符号：别处乱码丢失 ❌ —— $...$、$$...$$、\begin{equation} 等结构完整保护 ✅ 💻 Code blocks: mistranslated by others ❌ — smart-detected and auto-skipped, inline code included ✅ 💻 代码块：别处错误翻译 ❌ —— 智能检测自动跳过，行内代码也不误翻 ✅ 🔗 Links & formatting: usually stripped away ❌ — inline links, bold and emphasis survive translation intact ✅ 🔗 链接与格式：通常全部丢失 ❌ —— 行内链接、加粗、强调在译文中原样保留 ✅ 📖 Paper layout: usually disrupted ❌ — translation displayed below the original, side by side, webpage layout untouched ✅ 📖 论文排版：通常被打乱 ❌ —— 译文显示在原文下方原位对照，网页排版不变 ✅ 📕 5. PDF & Comic Translation (sign in, monthly free allowance) 📕 5. PDF 与漫画翻译（登录即享每月免费额度） 📄 PDF translation: upload a PDF, get back a fully re-typeset translated document — formulas and layout intact. Jobs run server-side and notify you when done, even if you close the page 📄 PDF 翻译：上传 PDF，得到完整重新排版的译文文档——公式与版式保持不变。任务在服务器端运行，关掉页面也会在完成时通知你 💬 Comic translation: speech bubbles detected, erased, and redrawn in your language; colorize mode included 💬 漫画翻译：自动识别气泡、擦除原文并以目标语言重绘；还支持黑白漫画上色模式 🎨 6. Your Translator, Your Style 🎨 6. 你的翻译，你的风格 ⚙️ Custom system prompt: academic rigor, casual tone, or domain terminology — you define it ⚙️ 自定义 System Prompt：学术严谨、口语表达、领域术语规范——随你定义 🌍 10+ UI languages; target language follows your browser by default 🌍 10+ 种界面语言；目标语言默认跟随浏览器设置 🌙 Dark mode for day-and-night reading 🌙 深色模式，昼夜皆宜 🚀 High-performance batch translation: concurrent workers, smart batching 🚀 高性能批量翻译：多并发 worker，智能分批 🎯 Use Cases 🎯 使用场景 🔬 Researchers: arXiv, IEEE, Nature papers with formulas intact; translate PDFs whole 🔬 科研工作者：arXiv、IEEE、Nature 论文公式无损；PDF 整篇翻译 👨‍💻 Developers: GitHub, Stack Overflow, technical docs — code stays code 👨‍💻 程序员：GitHub、Stack Overflow、技术文档——代码永远不被误翻 🎓 Students: foreign textbooks, lecture videos with bilingual subtitles 🎓 学生：外文教材、配双语字幕的课程视频 📊 Knowledge workers: foreign news, industry reports, replying in any language 📊 知识工作者：外媒新闻、行业报告，并用任意语言直接回复 📚 Comic & manga readers: translated speech bubbles, redrawn in place 📚 漫画爱好者：气泡原位翻译重绘 💡 In One Sentence 💡 一句话总结 🔑 Your Key, Your Data, Your Rules. 🔑 你的 Key，你的数据，你的规则。 From web pages to papers, videos, images, PDFs and comics — one translator for everything you read 🎓 从网页到论文、视频、图片、PDF 与漫画——一个翻译器，覆盖你阅读的一切 🎓

Details & SEO