Context Reader

The word you're stuck on, explained in your language — using the sentence around it. Free, open source, no tracking. **Understand the word, not just its definition.** A dictionary gives you every meaning of "culture". Context Reader reads the sentence you're actually looking at, works out that you're reading a lab report, and gives you the one that fits. Select a word on any page. You get the meaning that matches the context, in your language, in about two seconds. **110 languages** Every language Google lists Gemini as supporting, from Afrikaans to Zulu — සිංහල, தமிழ், हिन्दी, বাংলা, اردو, العربية, 简体中文, 繁體中文, 日本語, 한국어, Español, Français, Kiswahili and the rest. English and Sinhala are tuned with checked examples; the extension tells you which languages are and which aren't. **Already reading in your own language?** It won't translate a word into itself — it gives you a simpler word for the one you're stuck on. **Go deeper when you need to.** Three further views on any word: a worked scenario, a dictionary-style entry, and a plain explanation pitched at a thirteen-year-old. **Free, and built to stay that way.** Context Reader has no server. It calls Google's Gemini API directly from your browser using your own free API key, which takes about a minute to create. Because nobody is paying per-lookup costs, there is no subscription, no quota and no reason to ever start charging. **No tracking. Verifiably.** No analytics. No telemetry. No account. No identifiers. The extension holds permission for exactly one API — Google's Gemini — and your lookups go nowhere else. (Its typefaces ship inside the extension for every language except Chinese, Japanese and Korean; those three fetch one font file when a bubble opens, with no page address attached.) The entire source is public, so you don't have to take our word for it. Your API key is stored on your device, read only by the extension's background worker, and can be encrypted with a passphrase. **Open source (MIT).** Adding examples for your language is a few lines in one file. Source and privacy policy: https://github.com/contextreader/contextreader.github.io Questions: contextreader@gmail.com