Skip to content

Benchmark

Multi-SWE-bench Rust Verified

Coding benchmarks compare models on programming tasks. Scores are percentages. Higher is better.

Models measured
10
Scores recorded
30
Sources
1
IQ weight
2.64

Score distribution

5.7%
Mean
5%
Median
3.5%
Std dev
2.1%
10th percentile
10%
90th percentile
0.418%
Min
0.418%15.9%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.64 for chat models, measured over 10 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Claude Sonnet 3.7MopenHands, 20250329_MopenHands_Claude-3.7-Sonnet· Multi-SWE-bench Experiments data docs15.9%
2Gemini 2.5 ProMopenHands, 20250426_MopenHands_Gemini-2.5-Pro· Multi-SWE-bench Experiments data docs14.6%
3Claude 3.5 SonnetMopenHands, 20250329_MopenHands_Claude-3.5-Sonnet(Oct)· Multi-SWE-bench Experiments data docs12.1%
4Gemini 2.5 ProMSWE-agent, 20250426_MSWE-agent_Gemini-2.5-Pro· Multi-SWE-bench Experiments data docs10%
5o3 MiniMagentLess, 20250329_MagentLess_OpenAI-o3-mini-high· Multi-SWE-bench Experiments data docs7.9%
6o1MagentLess, 20250329_MagentLess_OpenAI-o1· Multi-SWE-bench Experiments data docs7.1%
7Claude 3.5 SonnetMSWE-agent, 20250329_MSWE-agent_Claude-3.5-Sonnet(Oct)· Multi-SWE-bench Experiments data docs6.7%
8Claude Sonnet 3.7MSWE-agent, 20250329_MSWE-agent_Claude-3.7-Sonnet· Multi-SWE-bench Experiments data docs6.7%
9DeepSeek R1MagentLess, 20250329_MagentLess_DeepSeek-R1· Multi-SWE-bench Experiments data docs6.7%
10Llama 4 MaverickMagentLess, 20250426_MagentLess_Llama-4-Maverick· Multi-SWE-bench Experiments data docs6.3%
11DeepSeek V3MSWE-agent, 20250329_MSWE-agent_DeepSeek-V3· Multi-SWE-bench Experiments data docs5.9%
12GPT 4oMagentLess, 20250329_MagentLess_GPT-4o-1120· Multi-SWE-bench Experiments data docs5.9%
13Claude Sonnet 3.7MagentLess, 20250329_MagentLess_Claude-3.7-Sonnet· Multi-SWE-bench Experiments data docs5.4%
14Gemini 2.5 ProMagentLess, 20250426_MagentLess_Gemini-2.5-Pro· Multi-SWE-bench Experiments data docs5.4%
15Claude 3.5 SonnetMagentLess, 20250329_MagentLess_Claude-3.5-Sonnet(Oct)· Multi-SWE-bench Experiments data docs5%
16DeepSeek V3MagentLess, 20250329_MagentLess_DeepSeek-V3· Multi-SWE-bench Experiments data docs5%
17Llama 4 MaverickMopenHands, 20250426_MopenHands_Llama-4-Maverick· Multi-SWE-bench Experiments data docs5%
18o3 MiniMopenHands, 20250329_MopenHands_OpenAI-o3-mini-high· Multi-SWE-bench Experiments data docs5%
19o3 MiniMSWE-agent, 20250329_MSWE-agent_OpenAI-o3-mini-high· Multi-SWE-bench Experiments data docs5%
20DeepSeek R1MopenHands, 20250329_MopenHands_DeepSeek-R1· Multi-SWE-bench Experiments data docs4.6%
21DeepSeek V3MopenHands, 20250329_MopenHands_DeepSeek-V3· Multi-SWE-bench Experiments data docs4.6%
22o1MSWE-agent, 20250329_MSWE-agent_OpenAI-o1· Multi-SWE-bench Experiments data docs4.2%
23GPT 4oMopenHands, 20250329_MopenHands_GPT-4o-1120· Multi-SWE-bench Experiments data docs3.3%
24Llama 4 MaverickMSWE-agent, 20250426_MSWE-agent_Llama-4-Maverick· Multi-SWE-bench Experiments data docs2.5%
25o1MopenHands, 20250329_MopenHands_OpenAI-o1· Multi-SWE-bench Experiments data docs2.5%
26Qwen2.5 72B InstructMagentLess, 20250329_MagentLess_Qwen2.5-72B-Instruct· Multi-SWE-bench Experiments data docs2.5%
27DeepSeek R1MSWE-agent, 20250329_MSWE-agent_DeepSeek-R1· Multi-SWE-bench Experiments data docs2.1%
28GPT 4oMSWE-agent, 20250329_MSWE-agent_GPT-4o-1120· Multi-SWE-bench Experiments data docs2.1%
29Qwen2.5 72B InstructMopenHands, 20250329_MopenHands_Qwen2.5-72B-Instruct· Multi-SWE-bench Experiments data docs1.7%
30Qwen2.5 72B InstructMSWE-agent, 20250329_MSWE-agent_Qwen2.5-72B-Instruct· Multi-SWE-bench Experiments data docs0.418%

Data credits: Multi-SWE-bench experiments (multi-swe-bench/experiments), Apache-2.0; resolved-rate fractions converted to percentages (licence)