Model-Context-Protocol server that integrates the Weblate translation platform with AI assistants (NestJS/TypeScript implementation).
https://github.com/mmntm/weblate-mcpStop context-switching between your IDE and Weblate's web interface. This MCP server puts your entire translation workflow at your fingertips through natural language commands.
You're deep in code when you need to check French translations for your frontend component. You open a browser tab, navigate to Weblate, click through projects, find the right component, filter by language, and finally locate the strings you need. By the time you're back to your editor, you've lost your train of thought.
The Weblate MCP server eliminates this friction entirely. Ask your AI assistant "Show me the French translations for the frontend component" and get immediate results without leaving your development environment.
This isn't another translation tool - it's your existing Weblate instance talking directly to your AI assistant. Full API coverage means you can:
The server handles authentication, API calls, and data formatting automatically. You focus on the translation work, not the tooling.
Pre-release Translation Audits Instead of manually checking each language in the Weblate interface, ask your AI assistant to generate a comprehensive translation status report across all components. Get incomplete translations, missing strings, and progress percentages in seconds.
CI/CD Integration Your deployment pipeline can query translation completeness before releases. No more shipping with incomplete localizations because someone forgot to check Weblate before the build.
Content Team Collaboration When your content team updates copy, they can ask the AI assistant to identify which translations need updates across all languages. The assistant understands the relationships between components and can provide actionable lists.
Quality Assurance Reviews Review translations contextually within your development workflow. Ask for specific translations, compare versions, and update content without breaking your review process.
Get started in under 60 seconds with npx - no installation required:
{
"mcpServers": {
"weblate": {
"command": "npx",
"args": ["-y", "@mmntm/weblate-mcp"],
"env": {
"WEBLATE_API_URL": "https://your-weblate-instance.com/api",
"WEBLATE_API_TOKEN": "your-api-token"
}
}
}
}
Works with Claude Desktop, Cursor, VS Code, and any MCP-compatible client. The server supports multiple transport methods (HTTP/SSE, STDIO) to fit your existing setup.
TypeScript Throughout: Full type safety with comprehensive error handling means fewer debugging sessions and better IntelliSense support.
Production Ready: Built on NestJS with proper dependency injection, request throttling, and input validation. This isn't a prototype - it's designed for real translation workflows.
Natural Language Interface: Ask questions the way you think about translations. The AI assistant understands context about projects, components, and languages without requiring specific API knowledge.
The comprehensive tool set covers every aspect of Weblate management through your AI assistant. No more switching between development tools and browser tabs to manage translations.
Whether you're using Claude Desktop for development, Cursor for AI-assisted coding, or VS Code with MCP extensions, this server adapts to your workflow. The HTTP transport option means web-based tools can integrate seamlessly.
The server runs locally or deploys to your infrastructure. Environment-based configuration means different team members can connect to different Weblate instances without conflicts.
Your translation workflow becomes part of your development process instead of a separate administrative task. That's the difference between managing translations and having translations managed for you.