GPT-5.6 Sol Beats Claude 5 Opus on ARC-AGI-3 Test via Custom Evaluation System
OpenAI claims GPT-5.6 Sol achieved a 38.3% score on the ARC-AGI-3 test, surpassing Claude 5 Opus. However, this milestone was only reached using the company's own custom evaluation system, prompting skepticism from industry observers.
