List your company

Fine-grained Visual Reasoning on V-Star, HR-Bench-4K, and HR-Bench-8K

85.9Average ScoreP2R-8B
58.96465.95772.9579.943Jul 1, 2026

Evaluation results

MethodLinks
P2R-8BParameters=8B2026.0785.9
P2R-4BParameters=4B2026.0785.2
Qwen3-VL-Instruct-32BParameters=32B2026.0779.7
DeepEyes-7BParameters=7B2026.0779.3
P2R-2BParameters=2B2026.0778.1
Thyme-7BParameters=7B2026.0777.1
ZoomEye-7BParameters=7B2026.0776.5
Qwen3-VL-Instruct-8BParameters=8B2026.0776.2
Qwen3-VL-Instruct-4BParameters=4B2026.0774.2
PixelReasoner-7BParameters=7B2026.0773.5
Qwen3-VL-Instruct-2BParameters=2B2026.0770
GPT-4o2026.0760