Skip to content

Benchmark

MCP-Mark

Agentic benchmarks measure how well models carry out tasks that take several steps. Scores are percentages.

Models measured
9
Scores recorded
9
Sources
1
IQ weight
0.46

Score distribution

51%
Mean
48.2%
Median
14.1%
Std dev
37%
10th percentile
60.8%
90th percentile
33.5%
Min
33.5%81.1%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.46 for chat models, measured over 9 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Kimi K2.7 Code· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim81.1%
2Qwen3.7 Max· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim60.8%
3Qwen3.7 Plus· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim58.7%
4Kimi K2.6· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim55.9%
5Qwen3.6 Plus· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim48.2%
6Qwen3.5 397B A17B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim46.1%
7DeepSeek V3.2· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim38%
8Qwen3.6 35B A3B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim37%
9Qwen3 Max· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim33.5%

Data credits: Data by LLM Stats