Benchmark
LiveCodeBench Pro Easy
Coding benchmarks compare models on programming tasks. Scores are percentages.
- Models measured
- 6
- Scores recorded
- 6
- Sources
- 1
- IQ weight
- 0.42
Score distribution
- 73.2%
- Mean
- 85.7%
- Median
- 16.5%
- Std dev
- 53.5%
- 10th percentile
- 90.1%
- 90th percentile
- 50.7%
- Min
This benchmark has a weight of 0.42 for chat models, measured over 6 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | GPT 5Easy, high· LiveCodeBench Pro data | 90.1% |
| 2 | GPT 5.2Easy· LiveCodeBench Pro data | 90.1% |
| 3 | o4 MiniEasy· LiveCodeBench Pro data | 85.7% |
| 4 | Qwen3 Next 80B A3BEasy, thinking· LiveCodeBench Pro data | 69% |
| 5 | Claude Sonnet 4.5Easy, thinking· LiveCodeBench Pro data | 53.5% |
| 6 | DeepSeek R1Easy· LiveCodeBench Pro data | 50.7% |