Google holds the highest implied probability at 24.5% to finish third on the Code Arena WebDev leaderboard by end of December, followed closely by Meta at 20%, Alibaba at 18.5%, OpenAI at 17.5%, and Moonshot at 17%. Current arena.ai lab rankings place Anthropic and OpenAI in the top two spots with models scoring above 1780, leaving Google, Alibaba, Moonshot, and Meta clustered within roughly 30 points for third. This tight grouping reflects comparable agentic coding performance on community-voted front-end tasks, where small gains from new releases or reasoning improvements can reorder the middle of the pack. Traders are pricing in continued rapid iteration across these labs through year-end, with limited differentiation in verified benchmarks to favor any single contender.