benchgap
Vision & documents

SimpleVQA leaderboard

As of 2026-11-21, the highest measured score on SimpleVQA is 81.7% by Qwen3.7 Plus. 64 more models have estimated scores, calibrated from the benchmarks they were measured on.

Measured scores: benchlm.ai.

#ModelScoreSource
1Qwen3.7 Plus81.7%measured
2Step 3.7 Flash79.2%measured
3Glimpse77.8%estimated ± 9.2 pp, low confidence
4Kimi K377.5%estimated ± 7.3 pp, medium confidence
5Inkling 2 Small76.2%estimated ± 7.3 pp, medium confidence
6Seed 2.1 Pro76.0%estimated ± 7.3 pp, medium confidence
7Qwen3.8-Omni-Flash75.2%estimated ± 7.3 pp, medium confidence
8GPT-6.1 Terra75.1%estimated ± 7.5 pp, low confidence
9Qwen3.8 Max75.0%measured
10Qwen3.8-Flash-Next74.1%estimated ± 7.3 pp, medium confidence
11Seed 2.1 Turbo73.6%estimated ± 7.3 pp, medium confidence
12Muse Spark Open73.6%measured
13Kimi K3.1 Pro73.5%estimated ± 7.5 pp, low confidence
14Gemini 3 Pro73.2%measured
15MiniMax M372.9%estimated ± 7.1 pp, low confidence
16Gemini 3.5 Flash72.9%estimated ± 7.5 pp, medium confidence
17Inkling 2 Mini72.6%estimated ± 7.3 pp, medium confidence
18dots3-note Preview72.5%measured
19Gemini 3.1 Pro72.4%measured
20GPT-5.6 Sol72.2%estimated ± 7.5 pp, medium confidence
21Qwen3.8-27B72.0%estimated ± 7.1 pp, low confidence
22Muse Spark71.3%measured
23Kimi K2.571.2%measured
24Kimi K2.671.1%estimated ± 7.3 pp, medium confidence
25GPT-5.570.1%estimated ± 7.5 pp, medium confidence
26GPT-5.6 Terra69.5%estimated ± 7.5 pp, medium confidence
27Qwen3.5 397B67.1%measured
28Hy467.0%estimated ± 7.5 pp, medium confidence
29GPT-5.6 Luna66.8%estimated ± 7.5 pp, medium confidence
30Grok 4.366.4%estimated ± 7.5 pp, medium confidence
31Pareto 26.966.3%estimated ± 7.5 pp, medium confidence
32MiMo-V2.566.2%estimated ± 7.5 pp, medium confidence
33Claude Opus 4.565.7%measured
34Claude Opus 4.665.5%estimated ± 7.5 pp, medium confidence
35Qwen3.6 Plus65.1%estimated ± 7.1 pp, medium confidence
36Gemma 4 31B65.0%estimated ± 7.5 pp, medium confidence
37GPT-5.4 mini64.6%estimated ± 7.5 pp, medium confidence
38Gemma 4 12B64.0%estimated ± 7.3 pp, medium confidence
39Step 5 Preview63.9%estimated ± 7.5 pp, medium confidence
40Hy4 flash63.7%estimated ± 7.5 pp, medium confidence
41Qwen3.5-122B-A10B63.4%estimated ± 7.1 pp, medium confidence
42Nemotron 3.5 Super VL 120B A12B62.6%estimated ± 7.3 pp, medium confidence
43Nemotron 3.5 Super VL 120B A12B NVFP462.2%estimated ± 7.3 pp, medium confidence
44Qwen3.5-27B62.1%estimated ± 7.1 pp, medium confidence
45Inkling-Small61.6%estimated ± 7.5 pp, medium confidence
46Muse Glimmer 30B61.6%estimated ± 7.5 pp, medium confidence
47Qwen3.5-35B-A3B61.4%estimated ± 7.1 pp, medium confidence
48Gemma 4 26B A4B61.3%estimated ± 7.5 pp, medium confidence
49GPT-5.461.1%measured
50Inkling61.0%estimated ± 7.5 pp, medium confidence
51Qwen3-VL-235B-A22B Instruct59.2%estimated ± 7.1 pp, medium confidence
52Qwen3.6-35B-A3B58.9%measured
53Interfaze Beta58.2%estimated ± 7.5 pp, medium confidence
54Grok 4.2057.4%measured
55Claude 4 Opus Thinking57.4%estimated ± 7.1 pp, medium confidence
56GLM-4.6V57.0%estimated ± 7.1 pp, medium confidence
57Ternary Bonsai 2 27B56.9%estimated ± 12.3 pp, low confidence
58Command A+56.2%estimated ± 7.1 pp, medium confidence
59Claude 3.7 Sonnet Thinking56.1%estimated ± 7.1 pp, medium confidence
60Qwen3.6-27B56.1%measured
61GPT-5.255.8%measured
62Claude 4 Opus55.1%estimated ± 7.1 pp, medium confidence
63Gemini 2.5 Flash-Lite54.4%estimated ± 7.1 pp, medium confidence
64Claude 3.7 Sonnet53.5%estimated ± 7.1 pp, medium confidence
65GLM-4.6V-Flash53.0%estimated ± 7.1 pp, medium confidence
66GPT-5.4 nano52.7%estimated ± 7.5 pp, medium confidence
67Nemotron 3 Nano Omni 30B A3B52.7%estimated ± 7.1 pp, medium confidence
68Gemini 2.0 Flash51.5%estimated ± 7.1 pp, medium confidence
69MiniMax-VL-0150.8%estimated ± 7.1 pp, medium confidence
70Grok 248.9%estimated ± 7.1 pp, medium confidence
71Gemini 2.0 Flash-Lite48.1%estimated ± 7.1 pp, medium confidence
72Mistral Small 3.147.2%estimated ± 7.1 pp, medium confidence
73Grok 2 mini46.5%estimated ± 7.1 pp, medium confidence
74Gemini 1.5 Flash42.5%estimated ± 7.1 pp, medium confidence
75Gemma 4 E4B41.5%estimated ± 7.5 pp, medium confidence
76Qwen2.5-VL-32B39.7%estimated ± 7.5 pp, medium confidence
77Gemma 4 E2B37.4%estimated ± 7.5 pp, medium confidence
78LFM3-VL-2B36.3%measured
79LFM2.5-VL-3B35.4%measured
80ZAYA1-VL-8B32.5%estimated ± 7.1 pp, low confidence
81LFM2.5-VL-450M21.6%estimated ± 7.1 pp, low confidence
82North Micro Vision Instruct19.8%estimated ± 12.3 pp, low confidence

All leaderboards

Agentic · terminal

Agentic · tools

Coding

Math

Knowledge & reasoning

Instruction following

Multilingual

Vision & documents

Long context

General