Instruction Following Benchmark leaderboard
16 ranked models · higher is better
View accessible chart data
| Model | Rank | Provider | Score |
|---|---|---|---|
| MAI-Thinking-1 | #1 | Microsoft | 85% |
| Nemotron 3 Ultra | #2 | NVIDIA | 81.7% |
| Grok 4.3 | #3 | xAI | 81.3% |
| Inkling | #4 | Thinking Machines Lab | 79.8% |
| Qwen3.7-Max | #5 | Alibaba Cloud | 79.1% |
| Qwen3.7-Plus | #6 | Alibaba Cloud | 79.1% |
| Gemini 3.5 Flash | #7 | 76.3% | |
| Qwen3.6 Plus | #8 | Alibaba Cloud | 75.8% |
| Nemotron 3 Nano Omni 30B A3B | #9 | NVIDIA | 74.2% |
| Hy3 Preview | #10 | Tencent | 63.1% |
| Claude Opus 4.5 | #11 | Anthropic | 58% |
| Ling 2.6 Flash | #12 | InclusionAI | 57% |
| LFM2.5-8B-A1B | #13 | LiquidAI | 56.5% |
| ZAYA1-8B | #14 | Zyphra | 52.6% |
| MiniCPM5-1B | #15 | OpenBMB | 46.7% |
| LFM2.5-230M | #16 | LiquidAI | 38.4% |