MCPMark-Verified: leaderboard
Moonshot AI reported human-verified edition of MCPMark for MCP tool-use tasks across Notion, GitHub, filesystem, Postgres, and Playwright environments.
Metric: Score (%). Source: huggingface.co. Source: benchmarklist.com. Saturation forecast: Around January 2027. 4 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.5 | 92.9 | #31 |
| 2 | Kimi K2.7 Code | 81.1 | #89 |
| 3 | Claude Opus 4.8 | 76.4 | #37 |
| 4 | Kimi K2.6 | 72.8 | #104 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=mcpmark-verified · How It Works · Data refreshed daily, snapshot 2026-10-20.