OpenAI holds the strongest market-implied odds at 57.5% to finish as the second-ranked lab on Code Arena’s WebDev leaderboard by end of December, reflecting its GPT-6 Astra Max model’s recent surge to within roughly 25–30 Elo points of Anthropic’s leading Claude Opus 5.5 Max on the community-voted arena.ai benchmark. Traders appear to price in OpenAI’s rapid iteration cadence and strong agentic coding results across frontend and full-stack tasks, even as Anthropic maintains the top lab position and draws only 25% odds for second place. A broad group of labs clustered near 50%—including Google, Meta, Alibaba, Moonshot, and several Chinese open-weight providers—highlights the tight competitive spread below the frontier, where incremental model releases, cost-performance shifts, and category-specific strengths in multi-step web workflows could reorder rankings before year-end. Upcoming flagship updates and continued community voting volume will likely determine whether OpenAI consolidates second or faces challenges from fast-improving contenders.