Share your thoughts, 1 month free Claude Pro on usSee more

Multimodal Reasoning on LVista

73.8Accuracy

Gemini-2.5-Pro

Updated 25d ago

Evaluation Results

Method	Links
Gemini-2.5-Pro 2026.06		73.8
GPT-5 2026.06		69.3
TACO (Ours-7B) 2026.06		55.6
Qwen3-VL-8B 2026.06		54.9
Qwen2.5-VL-32B 2026.06		54.4
GPT-4o 2026.06		53.2
PyVision-RL-7B 2026.06		49.2
Thyme-7B 2026.06		49
DeepEyes-v2-7B 2026.06		48.7
CodeV-RL-7B 2026.06		48.7
DeepEyes-7B 2026.06		47.7
Pixel-Reasoner-7B 2026.06		46.4
Mini-o3-7B 2026.06		45.7
InternVL3-8B 2026.06		45.6
MathCoder-VL-8B 2026.06		41.6
Qwen2.5-VL-7B 2026.06		39.8
LLaVA-OneVision-7B 2026.06		33.3