IFEval — Results

Rank	Model	Score
1	minicpm5-2b	86.7
2	lfm2-5-vl-3b	82.3
3	north-micro-vision-instruct	74.9
4	qwen3-5-27b	0.95
5	qwen3-7-plus	0.946
6	qwen3-6-plus	0.943
7	qwen3-7-max	0.943
8	o3-mini	0.939
9	qwen3-5-122b-a10b	0.934
10	qwen3-5-397b-a17b	0.926
11	llama-3-3-70b-instruct	0.921
12	qwen3-5-35b-a3b	0.919
13	qwen3-5-9b	0.915
14	kimi-k2-instruct-0905	0.898
15	kimi-k2-instruct	0.898
16	qwen3-5-4b	0.898
17	llama-3-1-nemotron-ultra-253b-v1	0.894
18	qwen3-235b-a22b-instruct-2507	0.887
19	llama-3-1-405b-instruct	0.886
20	gpt-4-1	0.874
21	deepseek-v3	0.861
22	phi-4-reasoning-plus	0.849
23	gpt-4-1-mini	0.841
24	qwen2-5-72b-instruct	0.841
25	qwq-32b	0.839

IFEval — Phases

1 phaseActive

Phase 1·

IFEval

Aug 3, 2026

Current