Skip to content

Benchmark

SWE-bench Verified

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

Models measured
130
Scores recorded
205
Sources
5
IQ weight
2.70

Score distribution

70%
Mean
73.5%
Median
13.7%
Std dev
51.6%
10th percentile
80.8%
90th percentile
8.7%
Min
8.7%96%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.70 for chat models, measured over 129 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
201Gemini 2.5 Flash-Lite· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim31.6%
202GPT-4.1 mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim23.6%
203DeepSeek V2.5· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim16.8%
204Nemotron 3.5 Lightning 30B A3Bagentic-coding figure · Codex v0.144.6 · unmodified system prompt · standard harness configuration· Reported by NVIDIA· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.11%
205GPT-4o mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim8.7%

Showing 201–205 of 205 results.

Data credits: Data by LLM Stats · Hugging Face official benchmark leaderboards; licence varies by benchmark · BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links. · Steel.dev Agent Leaderboards, MIT (licence) · OpenEvals/leaderboard-data, MIT (licence)