Anthropic’s Claude models, including recent releases like Claude Opus 5.5 and Fable 5.2 with expanded 1M-token context and adaptive reasoning, currently lead the Code Arena WebDev leaderboard by wide margins in agentic coding tasks involving multi-step tool use and frontend development. This performance edge drives the 77.5% implied probability, reflecting sustained dominance in practical web app benchmarks over competitors. OpenAI’s GPT-6 variants and Google’s Gemini models trail with lower scores, while smaller labs show limited upside. Traders appear focused on whether further Anthropic updates or rival drops before year-end could alter the December resolution.