Anthropic’s September 22, 2026 release of Claude Opus 5.5, which scored 54.6% on the newly introduced HLE-Diamond benchmark of 1,000 expert reasoning and knowledge questions, sets the baseline for expectations around the next Opus 5.6 or higher. OpenAI’s GPT-6 Astra currently leads the no-tools leaderboard at 59.9%, with Gemini 4 Argon close behind, creating competitive pressure for Anthropic to close the gap through gains in agentic capabilities and adaptive reasoning. Rapid prior cadence—from Opus 5 to 5.5 in roughly two months—combined with unconfirmed October routing reports of a newer checkpoint, suggests traders are pricing in a near-term flagship update that could meet or exceed the 55% threshold on debut. Key upcoming catalysts include any official model announcements, HLE refreshes, or expanded tool-use evaluations before year-end.
Experimental AI-generated summary referencing Polymarket data. This is not trading advice and plays no role in how this market resolves. · UpdatedView resolved

Beware of external links.
Beware of external links.
Frequently Asked Questions