ComplexMCP — leaderboard
Metric: Success Rate (self-reported). Source: benchmarklist.com. 17 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3 Flash (Preview) | 55.31 |
| 2 | GLM-4.7 | 42.55 |
| 3 | Claude Opus 4 | 41.84 |
| 4 | Claude Sonnet 4.5 | 39.71 |
| 5 | Claude Sonnet 4 | 38.29 |
| 6 | Llama 3.1 405B Instruct | 26.94 |
| 7 | Claude 3.5 Sonnet | 26.26 |
| 8 | Gemini 2.5 Pro | 24.81 |
| 9 | Llama 3.3 70B Instruct | 21.89 |
| 10 | DeepSeek V3 | 19.86 |
| 11 | GPT-5.1 | 19.14 |
| 12 | GPT-4o | 14.89 |
| 13 | Llama 3.1 8B Instruct | 8.51 |
Interactive version: theaggregate.ai/benchmark?slug=complexmcp · How the rankings work · Data refreshed daily, snapshot 2026-07-22.