Skip to content

Benchmark

Multi-SWE-bench Go Verified

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

Models measured
10
Scores recorded
30
Sources
1
IQ weight
1.45

Score distribution

4.3%
Mean
4%
Median
2.7%
Std dev
1.4%
10th percentile
6.8%
90th percentile
0%
Min
0%12.6%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.45 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
1Gemini 2.5 ProMopenHands, 20250426_MopenHands_Gemini-2.5-Pro· Multi-SWE-bench Experiments data docs12.6%
2Gemini 2.5 ProMSWE-agent, 20250426_MSWE-agent_Gemini-2.5-Pro· Multi-SWE-bench Experiments data docs9.8%
3Claude Sonnet 3.7MopenHands, 20250329_MopenHands_Claude-3.7-Sonnet· Multi-SWE-bench Experiments data docs7.5%
4Claude 3.5 SonnetMopenHands, 20250329_MopenHands_Claude-3.5-Sonnet(Oct)· Multi-SWE-bench Experiments data docs6.8%
5Gemini 2.5 ProMagentLess, 20250426_MagentLess_Gemini-2.5-Pro· Multi-SWE-bench Experiments data docs6.1%
6Claude 3.5 SonnetMSWE-agent, 20250329_MSWE-agent_Claude-3.5-Sonnet(Oct)· Multi-SWE-bench Experiments data docs5.8%
7Claude Sonnet 3.7MagentLess, 20250329_MagentLess_Claude-3.7-Sonnet· Multi-SWE-bench Experiments data docs5.8%
8Llama 4 MaverickMopenHands, 20250426_MopenHands_Llama-4-Maverick· Multi-SWE-bench Experiments data docs5.6%
9Claude Sonnet 3.7MSWE-agent, 20250329_MSWE-agent_Claude-3.7-Sonnet· Multi-SWE-bench Experiments data docs5.4%
10DeepSeek V3MagentLess, 20250329_MagentLess_DeepSeek-V3· Multi-SWE-bench Experiments data docs5.4%
11Claude 3.5 SonnetMagentLess, 20250329_MagentLess_Claude-3.5-Sonnet(Oct)· Multi-SWE-bench Experiments data docs5.1%
12o1MSWE-agent, 20250329_MSWE-agent_OpenAI-o1· Multi-SWE-bench Experiments data docs4.7%
13DeepSeek V3MSWE-agent, 20250329_MSWE-agent_DeepSeek-V3· Multi-SWE-bench Experiments data docs4.4%
14o1MagentLess, 20250329_MagentLess_OpenAI-o1· Multi-SWE-bench Experiments data docs4.4%
15o3 MiniMagentLess, 20250329_MagentLess_OpenAI-o3-mini-high· Multi-SWE-bench Experiments data docs4%
16o3 MiniMSWE-agent, 20250329_MSWE-agent_OpenAI-o3-mini-high· Multi-SWE-bench Experiments data docs4%
17DeepSeek R1MagentLess, 20250329_MagentLess_DeepSeek-R1· Multi-SWE-bench Experiments data docs3.7%
18o1MopenHands, 20250329_MopenHands_OpenAI-o1· Multi-SWE-bench Experiments data docs3.7%
19GPT 4oMopenHands, 20250329_MopenHands_GPT-4o-1120· Multi-SWE-bench Experiments data docs3.5%
20Llama 4 MaverickMagentLess, 20250426_MagentLess_Llama-4-Maverick· Multi-SWE-bench Experiments data docs3.5%
21GPT 4oMagentLess, 20250329_MagentLess_GPT-4o-1120· Multi-SWE-bench Experiments data docs2.8%
22Llama 4 MaverickMSWE-agent, 20250426_MSWE-agent_Llama-4-Maverick· Multi-SWE-bench Experiments data docs2.8%
23GPT 4oMSWE-agent, 20250329_MSWE-agent_GPT-4o-1120· Multi-SWE-bench Experiments data docs2.3%
24o3 MiniMopenHands, 20250329_MopenHands_OpenAI-o3-mini-high· Multi-SWE-bench Experiments data docs2.3%
25DeepSeek R1MSWE-agent, 20250329_MSWE-agent_DeepSeek-R1· Multi-SWE-bench Experiments data docs2.1%
26Qwen2.5 72B InstructMagentLess, 20250329_MagentLess_Qwen2.5-72B-Instruct· Multi-SWE-bench Experiments data docs1.4%
27Qwen2.5 72B InstructMopenHands, 20250329_MopenHands_Qwen2.5-72B-Instruct· Multi-SWE-bench Experiments data docs1.4%
28DeepSeek V3MopenHands, 20250329_MopenHands_DeepSeek-V3· Multi-SWE-bench Experiments data docs0.701%
29Qwen2.5 72B InstructMSWE-agent, 20250329_MSWE-agent_Qwen2.5-72B-Instruct· Multi-SWE-bench Experiments data docs0.467%
30DeepSeek R1MopenHands, 20250329_MopenHands_DeepSeek-R1· Multi-SWE-bench Experiments data docs0%

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