Benchmark
CyBench Guided
Coding benchmarks compare models on programming tasks. Scores are percentages.
- Models measured
- 9
- Scores recorded
- 9
- Sources
- 2
- IQ weight
- 0.43
Score distribution
- 11.6%
- Mean
- 12.5%
- Median
- 4.4%
- Std dev
- 5%
- 10th percentile
- 17.5%
- 90th percentile
- 5%
- Min
This benchmark has a weight of 0.43 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
| # | Model | Score |
|---|---|---|
| 1 | GPT 4o· CyBench official dataValues from 2 sources
| 17.5% |
| 2 | Claude 3.5 Sonnet· CyBench official dataValues from 2 sources
| 15% |
| 3 | Llama 3.1 405B Instruct· CyBench official dataValues from 2 sources
| 15% |
| 4 | Claude 3 Opus· CyBench official dataValues from 2 sources
| 12.5% |
| 5 | O1 Preview· CyBench official dataValues from 2 sources
| 10% |
| 6 | Llama 3 70B Instruct· Epoch AI Benchmarking Hub data | 7.5% |
| 7 | Gemini 1.5 Pro· CyBench official data | 5% |
| 8 | Mixtral 8x22B Instruct· CyBench official data | 5% |
| 9 | Mixtral 8x22B Instruct v0.1· Epoch AI Benchmarking Hub data | 5% |
Data credits: CyBench — cite Zhang et al. 2025 · Epoch AI Benchmarking Hub, CC BY 4.0 (licence)