shuji-bonji/xcomet-mcp-server

Overview

An MCP server that wraps xCOMET (eXplainable COMET) for translation quality evaluation. It exposes tools to score and explain machine translation outputs via the Model Context Protocol, enabling AI agents to assess translation quality programmatically.

Best for

Best for
Developers building translation evaluation tools or AI agents that need automated quality scoring

Use cases

• Evaluate machine translation quality in automated pipelines
• Compare translation outputs from different models or systems
• Integrate explainable translation scoring into AI agent workflows

Notes

An MCP server that wraps xCOMET (eXplainable COMET) for translation quality evaluation. It exposes tools to score and explain machine translation outputs via the Model Context Protocol, enabling AI agents to assess translation quality programmatically.

1 stars on GitHub. Last updated 2026-05-08. Licensed MIT.

Use cases

• Evaluate machine translation quality in automated pipelines
• Compare translation outputs from different models or systems
• Integrate explainable translation scoring into AI agent workflows

Pros

• Leverages state-of-the-art xCOMET for accurate, explainable evaluation
• Standard MCP interface simplifies integration with MCP-compatible clients
• Open source and written in TypeScript for broad developer accessibility

Cons

• Very early stage with only 1 star and minimal community adoption
• Requires local setup of xCOMET model dependencies, which may be resource-intensive
• Limited documentation and support due to niche status

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.