OpenAI’s GPT-6 Astra launched on September 3, 2026, yet failed to claim the top spot on the Text Arena leaderboard, where Anthropic’s Claude Fable 5 and Opus 4 variants currently hold the highest Elo scores above 1500. Trader consensus at 100% for “No” reflects Astra’s mid-pack placement—around 20th or lower in recent snapshots—behind multiple Claude and Gemini entries despite strong showings on specialized coding or agent benchmarks. This outcome aligns with historical patterns where new frontier models require iterative tuning before overtaking established leaders in blind human evaluations. Potential shifts could arise from rapid post-launch fine-tuning, revised Arena weighting, or undisclosed capability updates that close the gap before market resolution.