Anthropic's rapid iteration on frontier Claude models drives current trader sentiment, with Opus 5.5 and Fable 5.1 variants posting 61–67% on leading Humanity’s Last Exam leaderboards through enhanced reasoning chains, tool use, and targeted training on expert-level questions. These scores, verified across Artificial Analysis and provider harnesses as of early October 2026, already clear the 55% threshold while leaving meaningful headroom below expert human levels on the 2,500-question benchmark. Competitive pressure from OpenAI’s GPT-6 series and Google’s Gemini updates, alongside Anthropic’s ongoing releases, supports expectations for further gains before year-end. Key catalysts include any new model cards or benchmark updates demonstrating sustained progress in closed-book or tool-augmented settings.