Benchmark
HumanEval
Coding benchmarks compare models on programming tasks. Scores are percentages. Higher is better.
- Models and systems measured
- 98
- Scores recorded
- 99
- Sources
- 3
- IQ weight
- 1.81
Score distribution
- 71.5%
- Mean
- 78.7%
- Median
- 21.1%
- Std dev
- 37.8%
- 10th percentile
- 92%
- 90th percentile
- 11.6%
- Min
This benchmark has a weight of 1.81 for chat models, measured over 96 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model or system | Score |
|---|---|---|
| 1 | O1 Mini· EvalPlus Leaderboard dataValues from 2 sources
| 96.3% |
| 2 | O1 Preview· EvalPlus Leaderboard data | 96.3% |
| 3 | Kimi K2 0905· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 94.5% |
| 4 | GPT 5· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 93.4% |
| 5 | Kimi K2 Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 93.3% |
| 6 | GPT 4o· EvalPlus Leaderboard dataValues from 2 sources
| 92.7% |
| 7 | Qwen 2.5 Coder 32B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 92.7% |
| 8 | Sarvam-30B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 92.1% |
| 9 | Claude 3.5 Sonnet· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 92% |
| 10 | Mistral Large 2· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 92% |
| 11 | DeepSeek V3· EvalPlus Leaderboard data | 91.5% |
| 12 | DeepSeek V2.5· EvalPlus Leaderboard dataValues from 2 sources
| 90.2% |
| 13 | gemini-1.5-pro-002· EvalPlus Leaderboard data | 89% |
| 14 | Llama 3.1 405B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 89% |
| 15 | Nova Pro· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 89% |
| 16 | LongCat Flash Chat· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 88.4% |
| 17 | Mistral Small 3.1 24b Instruct 2503· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 88.4% |
| 18 | GPT 4· EvalPlus Leaderboard dataValues from 2 sources
| 88.4% |
| 19 | Grok 2· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 88.4% |
| 20 | Grok Beta· EvalPlus Leaderboard data | 88.4% |
| 21 | Llama 3.3 70B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 88.4% |
| 22 | Qwen2.5 32B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 88.4% |
| 23 | Qwen2.5 Coder 7B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 88.4% |
| 24 | Claude Haiku 3.5· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 88.1% |
| 25 | o1· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 88.1% |
| 26 | GPT-4.5 Preview· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 88% |
| 27 | Gemma 3 27B IT· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 87.8% |
| 28 | GPT-4o mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 87.2% |
| 29 | GPT-4 Turbo· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 87.1% |
| 30 | Qwen2.5 72B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 86.6% |
| 31 | Qwen2 72B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 86% |
| 32 | Grok 2 mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 85.7% |
| 33 | DeepSeek Coder V2 Instruct· EvalPlus Leaderboard data | 85.4% |
| 34 | Gemma 3 12B IT· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 85.4% |
| 35 | Nova Lite· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 85.4% |
| 36 | Claude 3 Opus· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 84.9% |
| 37 | Mistral Small 24B Instruct2501· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 84.8% |
| 38 | Qwen2.5 7B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 84.8% |
| 39 | Gemini 1.5 Pro· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 84.1% |
| 40 | Qwen 2.5 14B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 83.5% |
| 41 | Phi 4· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 82.6% |
| 42 | gemini-1.5-flash-002· EvalPlus Leaderboard data | 82.3% |
| 43 | DeepSeek Coder 33B Instruct· EvalPlus Leaderboard data | 81.1% |
| 44 | Nova Micro· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 81.1% |
| 45 | Llama 3.1 70B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 80.5% |
| 46 | Llama 3.1 70B Instructofficial instruction-tuned table· Reported by Meta· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 80.5% |
| 47 | Qwen2 7B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 79.9% |
| 48 | Qwen2.5-Omni 7B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 78.7% |
| 49 | Llama 3 70B Instruct· EvalPlus Leaderboard data | 77.4% |
| 50 | Claude 3 Haiku· EvalPlus Leaderboard dataValues from 2 sources
| 76.8% |
Showing 1–50 of 98 results.
Not ranked
The source marks these results as descriptive, not capability scores, so we list them by name instead of ranking them.
| Model or system | Score |
|---|---|
| DeepSeek V4.1 Flash Basepass@1 · 0-shot· Reported by DeepSeek· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 79.4% |
Data credits: EvalPlus Leaderboard and EvalPerf (evalplus.github.io), Apache-2.0; EvalPerf win-rate fractions converted to percentages (licence) · Data by LLM Stats · 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.