IFEval
OpenCompass · 48 models
| Rank | Model | Score |
|---|---|---|
| 1 | GPT-5-2025-08-07 (high) | 94.6 |
| 1 | JT-Think | 94.6 |
| 3 | TeleChat-thinking | 94.3 |
| 4 | Kimi-K2.5 | 93.9 |
| 5 | GLM-5 | 93.2 |
| 5 | Step-3.5-Flash | 93.2 |
| 7 | o4-mini-2025-04-16 (high) | 93 |
| 8 | GPT-5-mini-2025-08-07 (high) | 92.8 |
| 8 | Gemini-3-Pro-Preview | 92.8 |
| 10 | Kimi-K2-Thinking | 92.4 |
| 11 | o3-2025-04-16 | 92.2 |
| 12 | DeepSeek-V3.2-Speciale | 91.7 |
| 13 | Qwen3.5-397B-A17B | 91.5 |
| 14 | MiniMax-M2.5 | 91.1 |
| 15 | GLM-4.7 | 90.2 |
| 15 | GPT-5-nano-2025-08-07 (high) | 90.2 |
| 15 | LongCat-Flash-Chat | 90.2 |
| 15 | MiniMax-M2 | 90.2 |
| 15 | gpt-oss-120b (high) | 90.2 |
| 20 | Gemini-2.5-Pro | 90 |
| 21 | DeepSeek-V3.2 | 89.7 |
| 21 | Qwen3-30B-A3B-Thinking-2507 | 89.7 |
| 23 | MiMo-V2-Flash | 89.5 |
| 23 | Qwen3-Next-80B-A3B-Thinking | 89.5 |
| 25 | ERNIE-4.5-300B-A47B-PT | 88.9 |
| 25 | gpt-oss-20b (high) | 88.9 |
| 27 | Finix-P1-32B (Thinking) | 88.7 |
| 27 | GLM-4.6 | 88.7 |
| 29 | Kimi-K2-Instruct | 88.5 |
| 29 | Qwen3-4B-Thinking-2507 | 88.5 |
| 31 | Claude Sonnet 4 (Thinking) | 88.3 |
| 31 | Qwen3-235B-A22B-Instruct-2507 | 88.3 |
| 33 | GPT-4.1-20250414 | 88.2 |
| 34 | Qwen3-235B-A22B-Thinking-2507 | 87.8 |
| 35 | Qwen3-Next-80B-A3B-Instruct | 87.6 |
| 36 | Llama4-Maverick-17B-128E-Instruct | 87.1 |
| 37 | Qwen3-32B (Thinking) | 86 |
| 38 | Qwen3-8B (Thinking) | 85.6 |
| 39 | GLM-4.5 | 85.4 |
| 40 | Qwen3-30B-A3B-Instruct-2507 | 83.9 |
| 41 | Seed-OSS-36B-Instruct | 83.7 |
| 42 | Qwen3-4B-Instruct-2507 | 82.4 |
| 43 | ERNIE-4.5-21B-A3B-Thinking | 81.2 |
| 44 | Gemma-3-27B-it | 81 |
| 45 | DeepSeek-R1-0528 | 80 |
| 46 | doubao-seed-1.6-thinking-250615 | 78.9 |
| 47 | MiniCPM-SALA | 76.9 |
| 48 | Hunyuan-A13B-Instruct | 60.3 |