Benchmark
Multi-Challenge
Instruction following benchmarks measure how closely models follow instructions. Scores are percentages. Higher is better.
- Models measured
- 30
- Scores recorded
- 33
- Sources
- 2
- IQ weight
- 1.22
Score distribution
- 52.6%
- Mean
- 55.2%
- Median
- 14.6%
- Std dev
- 35.8%
- 10th percentile
- 67.6%
- 90th percentile
- 15%
- Min
This benchmark has a weight of 1.22 for chat models, measured over 30 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | Nova 2 Lite· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 76.6% |
| 2 | Nova 2 LiteAmazon Nova 2 Lite official launch chart / technical report · accuracy, v1· Reported by Amazon Web Services, Amazon AGI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 76.6% |
| 3 | GPT 5· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 69.6% |
| 4 | Qwen3.5 397B A17B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 67.6% |
| 5 | Qwen3.5 397B A17BQwen3.5 release/model card · published setting· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 67.6% |
| 6 | Doubao Seed 1.8· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 66.7% |
| 7 | Nemotron 3 Ultra· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 63.8% |
| 8 | Qwen3 Max· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 63.3% |
| 9 | Qwen3.5 122B-A10B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 61.5% |
| 10 | Seed 2.0 Mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 61.1% |
| 11 | Qwen3.5 27B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 60.8% |
| 12 | Step 3.5 FlashStep 3.5 technical report · Table 5 · PaCoRe· Reported by StepFun· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 60.8% |
| 13 | OpenAI o3· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 60.4% |
| 14 | Mercury 3· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 60.4% |
| 15 | Qwen3.5 35B A3B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 60% |
| 16 | Step 3.5 FlashStep 3.5 technical report · Table 5 · vanilla· Reported by StepFun· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 55.7% |
| 17 | Nemotron 3 Super 120B A12B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 55.2% |
| 18 | Qwen3.5 9B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 54.5% |
| 19 | Kimi K2 Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 54.1% |
| 20 | Mai Thinking 1· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 53% |
| 21 | Qwen3.5 4B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 49% |
| 22 | MiniMax M1 40K· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. | 44.7% |
| 23 | MiniMax M1 80K· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. | 44.7% |
| 24 | GPT-4.5 Preview· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 43.8% |
| 25 | o4 Mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 43% |
| 26 | GPT 4o· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 40.3% |
| 27 | o3 Mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 39.9% |
| 28 | Nemotron 3 Nano 30B A3B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 38.5% |
| 29 | GPT 4.1· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 38.3% |
| 30 | GPT-4.1 mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 35.8% |
| 31 | Qwen3.5 2B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 33.7% |
| 32 | Qwen3.5 0.8B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 18.9% |
| 33 | GPT-4.1 nano· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 15% |
Data credits: Data by LLM Stats · BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links.