AI model rankings
Understand capability differences before choosing AI resources.
EvalPlus · Coding ability
Historical benchmark · Source not updated for over 60 days; not a current model ranking
HumanEval+ · pass@1 · 版本 0.1.10 · Scores range from 0–100; higher is better. Missing scores are excluded, and ties share a rank. Search preserves original ranks.
| Rank | Model and version | Score | Evidence type |
|---|---|---|---|
| 1 | O1 Mini (Sept 2024)对话提示Model details ↗ | 89.0 | Leaderboard listing |
| 1 | O1 Preview (Sept 2024)对话提示Model details ↗ | 89.0 | Leaderboard listing |
| 3 | GPT 4o (Aug 2024)对话提示Model details ↗ | 87.2 | Leaderboard listing |
| 3 | Qwen2.5-Coder-32B-Instruct对话提示Model details ↗ | 87.2 | Leaderboard listing |
| 5 | DeepSeek-V3 (Nov 2024)对话提示Model details ↗ | 86.6 | Leaderboard listing |
| 5 | GPT-4-Turbo (April 2024)对话提示Model details ↗ | 86.6 | Leaderboard listing |
| 7 | DeepSeek-V2.5 (Nov 2024)对话提示Model details ↗ | 83.5 | Leaderboard listing |
| 7 | GPT 4o Mini (July 2024)对话提示Model details ↗ | 83.5 | Leaderboard listing |
| 9 | DeepSeek-Coder-V2-Instruct对话提示Model details ↗ | 82.3 | Leaderboard listing |
| 10 | Claude Sonnet 3.5 (June 2024)对话提示Model details ↗ | 81.7 | Leaderboard listing |
| 10 | GPT-4-Turbo (Nov 2023)对话提示Model details ↗ | 81.7 | Leaderboard listing |
| 12 | Grok Beta对话提示Model details ↗ | 80.5 | Leaderboard listing |
| 13 | GPT-4 (May 2023)对话提示Model details ↗ | 79.3 | Leaderboard listing |
| 13 | Gemini 1.5 Pro 002对话提示Model details ↗ | 79.3 | Leaderboard listing |
| 15 | CodeQwen1.5-7B-Chat对话提示Model details ↗ | 78.7 | Leaderboard listing |
| 16 | OpenCoder-8B-Instruct对话提示Model details ↗ | 77.4 | Leaderboard listing |
| 16 | claude-3-opus (Mar 2024)对话提示Model details ↗ | 77.4 | Leaderboard listing |
| 18 | Gemini 1.5 Flash 002对话提示Model details ↗ | 75.6 | Leaderboard listing |
| 19 | DeepSeek-Coder-33B-instruct对话提示Model details ↗ | 75.0 | Leaderboard listing |
| 20 | Codestral-22B-v0.1对话提示Model details ↗ | 73.8 | Leaderboard listing |
| 20 | OpenCodeInterpreter-DS-33B对话提示Model details ↗ | 73.8 | Leaderboard listing |
| 22 | WizardCoder-33B-V1.1对话提示Model details ↗ | 73.2 | Leaderboard listing |
| 23 | Artigenz-Coder-DS-6.7B对话提示Model details ↗ | 72.6 | Leaderboard listing |
| 24 | Llama3-70B-instruct对话提示Model details ↗ | 72.0 | Leaderboard listing |
| 24 | Mixtral-8x22B-Instruct-v0.1对话提示Model details ↗ | 72.0 | Leaderboard listing |
Source evidence and evaluation methodology
Source:EvalPlus · Coding ability · Apache-2.0. We organize original scores without creating cross-benchmark composite scores. Listing a result does not mean we independently reproduced it.
Attribution: EvalPlus contributors. Scores are compiled from EvalPlus. The original project uses Apache License 2.0; its license and notices are retained. Data is provided as-is without warranties. License and attribution for this source version
Source snapshot:9d8bdc12001f459cb5b3e8642f6a2f4f · Source commit：4ebfa0d4b62edb942232c723b9d1641fe1d24780 · Original content SHA-256:e6837dc3b94f32cb1da0ced012e52edc4a722007b684df5b63da029bbd3b46d2
Fetched on demand and published after admin review, not in real time. Undisclosed evaluation versions are not inferred.EvalPlus uses greedy-decoding pass@1; Different dimensions and test settings cannot be compared directly.