Skip to content

Benchmark

Multi-IF

Instruction following benchmarks measure how closely models follow instructions. Scores are percentages.

Models measured
20
Scores recorded
22
Sources
1
IQ weight
1.22

Score distribution

74%
Mean
75.1%
Median
6.6%
Std dev
66.1%
10th percentile
80.1%
90th percentile
57.2%
Min
57.2%87.7%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.22 for chat models, measured over 20 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Ling 3.0 Flash· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim87.7%
2Qwen 3 235B A22B Thinking 2507· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim80.6%
3LFM2.5 2.6B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim80.1%
4o3 Mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim79.5%
5Qwen3 VL 235Bthinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.79.1%
6Qwen3 VL 32B Thinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.78%
7Qwen3 Next 80B A3Bthinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim77.8%
8Qwen3 235B A22B Instruct 2507· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim77.5%
9Qwen3 VL 235B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.76.3%
10Qwen3 Next 80B A3B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim75.8%
11Qwen: Qwen3 VL 8B Thinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.75.1%
12Qwen3 VL 8B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.75.1%
13Qwen3 14B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim74.8%
14qwen3-vl-30bthinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.73%
15Qwen3 30B A3B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim72.2%
16Qwen3 VL 32B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.72%
17GPT 4.1· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim70.8%
18GPT-4.5 Preview· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim70.8%
19GPT-4.1 mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim67%
20qwen3-vl-30b· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.66.1%
21GPT 4o· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim60.9%
22GPT-4.1 nano· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim57.2%

Data credits: Data by LLM Stats