OpenAI's GPT-5.5 has claimed the top spot on DesignArena's Large Model Game Development leaderboard, achieving an Elo score of 1362. This surpasses Claude Opus 4.7, which holds a score of 1352. GPT-5.5's success is attributed to its verbose approach to writing frontend code, which has proven advantageous in producing fully functional games. This development highlights OpenAI's consistent edge over Anthropic in the game development sector.