OpenAI's GPT-6.1 Sol Scores Close to Flagship GPT-6 Astra on Benchmark

30 Sept 2026, 12:32 am · 2d ago · 1 min read · OfficeChai

OpenAI's newly released AI model, GPT-6.1 Sol, has scored just one point behind the company's flagship model, GPT-6 Astra, on the Artificial Analysis Intelligence Index. The benchmark evaluated the model across key performance criteria, including reasoning, coding capabilities, latency, and overall output quality. The result highlights that OpenAI is closing the capability gap between its high-efficiency, cost-effective models and its most powerful flagship offerings. This performance parity allows developers and businesses to deploy high-tier intelligence for enterprise workflows and consumer applications without incurring the higher latency and inference costs associated with top-tier flagship frontier models.