AA IFBench: leaderboard
Artificial Analysis independent evaluation of IFBench: instruction-following on 58 out-of-domain constraints (294 questions).
Metric: Accuracy (%). Source: artificialanalysis.ai. Saturation forecast: Around January 2027. 450 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Grok 4.3 (Medium) | 83.33 | #188 (Grok 4.3) |
| 2 | Grok 4.20 0309 (Reasoning) | 82.93 | #112 (Grok 4.20 0309) |
| 3 | MiniMax-M3 | 82.86 | #136 |
| 4 | Grok 4.3 (High) | 81.29 | #188 (Grok 4.3) |
| 5 | Grok 4.3 (Low) | 80.95 | #188 (Grok 4.3) |
| 6 | Qwen 3.7 Max | 80.54 | #75 |
| 7 | Nemotron Cascade 2 30B A3B | 80.41 | #551 |
| 8 | MiMo-V2.5-Pro (Reasoning) | 79.86 | #158 (MiMo-V2.5-Pro) |
| 9 | Nova 2.0 Pro Preview (Low) | 79.59 | #357 (Nova 2.0 Pro Preview) |
| 10 | Nova 2.0 Pro Preview (Medium) | 79.05 | #357 (Nova 2.0 Pro Preview) |
| 11 | Qwen 3.5 397B A17B (Reasoning) | 78.78 | #146 (Qwen 3.5 397B A17B) |
| 12 | Qwen 3.7 Plus | 77.96 | #92 |
| 13 | GPT-5.2 Codex (xHigh) | 77.62 | #95 (GPT-5.2 Codex) |
| 14 | Gemini 3.1 Flash Lite | 77.21 | #191 |
| 15 | Gemini 3.1 Pro (Preview) | 77.14 | #58 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=aa-ifbench · How It Works · Data refreshed daily, snapshot 2026-10-20.