Back to Benchmarks
📋
IFEval
Instruction Following Eval
500+ programmatically verifiable instruction constraints
📋 Instruction Following
Best: 94%
11 Models Evaluated
Champion
Claude Opus 4.7
Anthropic
94%
Best Score
Model Rankings
All models ranked by score on IFEval. Scores are relative to the best score of 94%.
| # | Model | Provider | Score | vs Best |
|---|---|---|---|---|
| 1 | Claude Opus 4.7 Elo 1500 | Anthropic | 94% | Best |
| 2 | Claude Opus 4.6 Elo 1503 | Anthropic | 94% | +0.0% |
| 3 | Kimi K2.5 Elo 1438 | Moonshot | 94% | +0.0% |
| 4 | Qwen 3.5 Elo 1450 | Alibaba | 92.6% | -1.4% |
| 5 | GPT-5.5 Elo 1478 | OpenAI | 91% | -3.0% |
| 6 | Nemotron Ultra 253B Elo 1348 | NVIDIA | 89.5% | -4.5% |
| 7 | GLM-5 Elo 1454 | Zhipu AI | 88% | -6.0% |
| 8 | MiniMax M2.5 Elo 1404 | MiniMax | 87.5% | -6.5% |
| 9 | Gemini 3.1 Pro Elo 1492 | 85% | -9.0% | |
| 10 | DeepSeek R1 Elo 1398 | DeepSeek | 83.3% | -10.7% |
| 11 | Gemini 3 Pro Elo 1486 | 82% | -12.0% |
Score Distribution
Claude Opus 4.7
94%
Claude Opus 4.6
94%
Kimi K2.5
94%
Qwen 3.5
92.6%
GPT-5.5
91%
Nemotron Ultra 253B
89.5%
GLM-5
88%
MiniMax M2.5
87.5%
Gemini 3.1 Pro
85%
DeepSeek R1
83.3%
Run This Benchmark
Learn how to evaluate your own models on IFEval.
IFEval is a standardized benchmark for evaluating AI model capabilities.
To run this benchmark on your own models, visit the official benchmark page or leaderboard for detailed instructions, code, and evaluation scripts.
Data compiled from official model cards, public leaderboards, and independent evaluations.