OpenAI’s GPT-6 Astra Max currently sits in second place on Arena.ai’s Code Arena WebDev leaderboard behind Anthropic’s Claude Opus 5.5 Max, with community preference votes on agentic frontend and full-stack tasks creating a tight margin under 10 Elo points as of early October 2026. This positioning, alongside Anthropic’s additional strength via Claude Sonnet 5.5, underpins OpenAI’s 39.5% market-implied odds for second-best lab by year-end. Moonshot’s earlier Kimi K3 surge and strong open-weight performance keep it at 13.9%, while Google, Alibaba, and Meta trail amid ongoing model iteration. New releases or benchmark shifts before December could still reorder the crowded frontier.