MCPEvol-Bench: leaderboard
MCPEvol-Bench (2026) tests LLM agents on 201 multi-server MCP tool-use tasks as the servers evolve through simulated developer updates. The Evolutionary Competency Score is the per-task mean minus the standard deviation of the judge-rated Task Fulfillment (1 to 10) across three server versions; higher is better.
Metric: Evolutionary Competency Score (per task, mean minus standard deviation of the Task Fulfillment score across the three server versions, averaged over tasks; about -1.2 to 10; 201 synthesized multi-server tasks over a fixed set of candidate MCP servers from 123 real servers; rubric scores on a 1 to 10 scale from a DeepSeek-Chat judge over the instruction, final answer and execution trajectory (MCP-Bench evaluation framework)). Source: arxiv.org. Saturation forecast: Around March 2027. 12 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 4.6 | 6.09 | #102 |
| 2 | Claude Sonnet 4.6 | 5.22 | #161 |
| 3 | GPT-5.4 | 5.2 | #123 |
| 4 | Claude Sonnet 4.5 (Thinking) | 4.52 | #250 (Claude Sonnet 4.5) |
| 5 | Gemma 4 31B (IT) | 4.45 | #342 |
| 6 | GPT-5.1 | 4.32 | #229 |
| 7 | O3 | 4.24 | #228 |
| 8 | Gemini 2.5 Pro | 3.84 | #266 |
| 9 | GPT-4o | 3.8 | #741 |
| 10 | Qwen 3.5 27B | 3.78 | #359 |
| 11 | Llama 3.3 70B | 3.56 | #1498 |
| 12 | Qwen 3.5 9B | 3.2 | #803 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #5 Claude Opus 5.5.
Interactive version: theaggregate.ai/benchmark?slug=mcpevol-bench · How It Works · Data refreshed daily, snapshot 2026-11-13.