LLM Stats (MCP-Mark): leaderboard
MCP tool-use benchmark measuring model ability to discover, invoke, and chain Model Context Protocol tools for multi-step task completion.
Metric: Score (%). Source: llm-stats.com. Source: qwen.ai. Saturation forecast: Around January 2027. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Kimi K2.7 Code | 81.1 | #89 |
| 2 | Qwen 3.7 Max | 60.8 | #75 |
| 3 | Qwen 3.7 Plus | 58.7 | #92 |
| 4 | GLM-5.1 (Thinking) | 57.5 | #97 (GLM-5.1) |
| 5 | DeepSeek V4 Pro (Reasoning, Max Effort) | 57.1 | #102 (DeepSeek V4 Pro) |
| 6 | Claude Opus 4.6 (Thinking) | 56.7 | #63 (Claude Opus 4.6) |
| 7 | Kimi K2.6 (Thinking) | 55.9 | #104 (Kimi K2.6) |
| 8 | Qwen 3.6 Plus | 48.2 | #120 |
| 9 | Qwen 3.5 397B A17B | 46.1 | #146 |
| 10 | DeepSeek V3.2 | 38 | #205 |
| 11 | Qwen 3.6 35B A3B | 37 | #226 |
| 12 | Qwen 3 Max (Thinking) | 33.5 | #208 (Qwen 3 Max) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=llm-stats-mcp-mark · How It Works · Data refreshed daily, snapshot 2026-10-20.