MCPEvol-Bench - Planning Effectiveness (Late Stage): leaderboard

Metric: Planning Effectiveness score (1 to 10 rubric; servers after five evolution rounds; 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 2028. 12 models tracked.

Top models

#ModelScore
1Gemma 4 31B (IT)6.68
2Claude Sonnet 4.66.17
3Claude Opus 4.66.03
4Qwen 3.5 27B5.12
5Claude Sonnet 4.5 (Thinking)4.98
6Qwen 3.5 9B4.86
7Llama 3.3 70B3.94
8GPT-5.43.87
9Gemini 2.5 Pro3.71
10GPT-5.13.5
11O33.1
12GPT-4o2.78

Interactive version: theaggregate.ai/benchmark?slug=mcpevol-bench-planning-effectiveness-late-stage · How It Works · Data refreshed daily, snapshot 2026-09-29.