List your company

Multimodal mathematical reasoning on MathVision

61.5Pass@1 AccuracyGPT-5
19.48430.39241.352.208Feb 18, 2026Mar 5, 2026Mar 20, 2026Apr 5, 2026Apr 20, 2026May 5, 2026May 21, 2026

Evaluation results

MethodLinks
GPT-5Model Type=Closed-Source2026.0561.5
GPT-5-Nano-HighModel Category=Closed-source Models2026.0258.75
Qwen3-VL-8B-ThinkingSeries=Qwen3-VL-8B2026.0257.89
Qwen3-VL-8B-DeepVisionSeries=Qwen3-VL-8B, Training Algorithm=GSPO2026.0255.49
MiMo-VL-7B-DeepVisionSeries=MiMo-VL-7B, Training Algorithm=GSPO2026.0255.24
MiMo-VL-7B-RL-2508Series=MiMo-VL-7B2026.0253.91
Kimi-K2.5Model Type=Open-Source & Baselines2026.0553.3
MiMo-VL-7B-OpenMMReasonerSeries=MiMo-VL-7B2026.0252.97
Gemini-2.5-Flash-LiteModel Category=Closed-source Models2026.0252.47
Qwen3-VL-8B-InstructSeries=Qwen3-VL-8B2026.0251.44
MiMo-VL-7B-MathBookSeries=MiMo-VL-7B2026.0251.31
MiMo-VL-7B-SFT-2508Series=MiMo-VL-7B2026.0250.69
MiMo-VL-7B-MM-EurekaSeries=MiMo-VL-7B2026.0250
Qwen3-VL-32BModel Type=Open-Source & Baselines2026.0545.4
MAESTROModel Type=Ours2026.0543.4
Gemini-2.5-FlashModel Type=Closed-Source2026.0539.8
Gemini-2.5-ProModel Type=Closed-Source2026.0539.8
GLM-4.6VModel Type=Open-Source & Baselines2026.0539.1
GPT-4oModel Type=Closed-Source2026.0530.4
VTOOL-R1Model Type=Think with Images Methods2026.0529.3
Untrained ModelModel Type=Open-Source & Baselines2026.0529
DeepEyes-v2Model Type=Think with Images Methods2026.0528.9
ThymeModel Type=Think with Images Methods2026.0527.6
VTS-VModel Type=Think with Images Methods2026.0527
DeepEyesModel Type=Think with Images Methods2026.0526.6
MathCoder-VLModel Type=Think with Images Methods2026.0526
Direct AnsweringModel Type=Open-Source & Baselines2026.0524.9
PixelReasonerModel Type=Think with Images Methods2026.0523.4
VisionReasonerModel Type=Think with Images Methods2026.0521.7
Visual-ARFTModel Type=Think with Images Methods2026.0521.4
Chain-of-FocusModel Type=Think with Images Methods2026.0521.1