OpenAI’s recent GPT-6 series releases, including GPT-6 Astra’s strong HLE results above 50% (and 82.9% on tool-assisted subsets) and the September 29 launch of cost-efficient GPT-6.1 Sol, have established a clear trajectory of frontier performance on the expert-level benchmark. This has driven trader consensus toward a high implied probability that the next Sol-class model (6.1+) will clear the 40% threshold at debut. Anthropic’s Claude variants currently lead public leaderboards near 65%, underscoring competitive pressure on OpenAI to showcase HLE scores early. Key upcoming catalysts include any DevDay announcements, HLE-Rolling updates, or verified benchmark publications that could confirm or delay resolution by mid-2027.