Standby Translator
Extension Actions
Privacy-focused translation using local Ollama LLM models
Standby Translator is a privacy-focused translation extension for people who want local AI translation while browsing.
Select text on a web page, right-click, and choose "Translate with Standby" to see a clean translation popup near the selection. You can also click the extension icon to paste text directly into a quick translation panel.
Key features:
- Local translation through your own Ollama endpoint
- Right-click translation for selected page text
- Quick paste translation from the extension icon
- Separate word and sentence prompts
- Word details including translation, part of speech, definition, pronunciation, and sample usage when your model provides them
- Clean sentence translations without extra clutter
- Force re-translate when you want a fresh result from a different or stronger model
- Local translation cache for faster repeated lookups
- Recent single-word cache view and Markdown word-card export
- Configurable source/target languages, prompts, model, timeout, and sampling parameters
Privacy notes:
Standby Translator does not use analytics, ads, or developer-controlled translation servers. Text is sent only to the Ollama endpoint you configure. Settings, prompts, and cached translations are stored locally in Chrome storage.
Setup note:
You need Ollama running locally. For browser-extension requests, start Ollama with CORS enabled:
```bash
OLLAMA_ORIGINS="*" ollama serve
```
Then install a model, for example:
```bash
ollama pull llama3
```