List your company

General AI Assistant Tasks on GAIA Level 2 (val)

54.2AccuracyGPT-5
21.9630.3338.747.07Dec 7, 2025

Evaluation results

MethodLinks
GPT-5Model=GPT-5, Web search=true, Trials per question=642025.1254.2—0.1870.745
GPT-4oModel=GPT-4o, Web search=true, Trials per question=642025.1223.2—0.1190.662