Anthropic’s April 2026 release of Claude Opus 4.7 and the Mythos Preview has established frontier-leading performance on Humanity’s Last Exam, a 2,500-question benchmark spanning advanced math, science, and humanities, with recent leaderboards showing scores from 36% to as high as 64.7% under tool-augmented and extended-reasoning conditions. These gains stem from iterative improvements in scalable oversight, larger context windows, and agentic capabilities that outpace OpenAI’s GPT-5 series and Google’s Gemini models on the same evaluation. Traders are focused on whether a public Claude variant sustains or exceeds the 30–35% threshold by the June 30 resolution date, with potential catalysts including developer previews, benchmark updates, or competitive releases at upcoming AI conferences that could shift the aggregated market-implied odds.