List your company

Physical Reasoning on CosmosReason1-Bench

64.7Overall ScoreGemini-2.5-Pro
49.82853.68957.5561.411Feb 11, 2026

Evaluation results

MethodLinks
Gemini-2.5-ProAccess=Closed-Source2026.0264.7458483.64574
PhyCritic-7BAccess=Open-Source, Parameters=7B2026.0263.9536895.53763
Cosmos-R1-7BAccess=Open-Source, Parameters=7B2026.026349.46383.858.860
Gemini-2.5-FlashAccess=Closed-Source2026.0257.942.768.278.235.562.8
GPT-4oAccess=Closed-Source2026.0256.3386478.23366
Eagle-2.5-8BAccess=Open-Source, Parameters=8B2026.0254.3396767.42968
Qwen2.5-VL-7BAccess=Open-Source, Parameters=7B2026.0254.340.45082.53857.6
Robobrain-2.0-7BAccess=Open-Source, Parameters=7B2026.0250.4374185.42956