Translate code, errors, and documentation directly inside VSCode.
🚧 Early development stage
Frilingo helps developers understand code, error messages, comments, documentation, and technical content in their preferred language using local LLMs through Ollama.
Translate the current editor selection using a local Ollama model.
Quickly translate documentation, comments, and annotations without leaving the editor.
Use local models through Ollama.
Examples:
- gemma3:4b
- qwen3:4b
- llama3.1
Manage local models directly from VS Code.
- Start Ollama
- List installed models
- Pull recommended models
- Check runtime status
Download and install Ollama:
Example:
ollama pull gemma3:4bInstall dependencies:
npm installRun type checking:
npm run check-typesRun lint:
npm run lintRun smoke test:
npm run test:ollama- Ollama integration
- Selection translation
- Model management
- Hover translation
- Inline replacement
- Multiple language support
- Translation caching
- User settings
Some reasoning-oriented models may generate explanations instead of direct translations.
For translation-focused workflows, models such as gemma3 are recommended.
MIT