Will GPT-6 Astra beat Claude on the Artificial Analysis coding benchmark before December 31, 2026?

Market context

GPT-6 Astra (OpenAI, September 3, 2026) and the Claude models (Anthropic) compete in the frontier AI model market. Artificial Analysis is the reference benchmark tracker. OpenAI claimed that Astra is 'the smartest model in the world' with a record on FrontierMath Tier 4.

Resolution criteria

• How it resolvesResolves YES if, in the last Coding Index update from Artificial Analysis published on or before December 31, 2026 at 23:59 Mexico City time, GPT-6 Astra's score is STRICTLY HIGHER than that of the top-ranked Anthropic (Claude) model in that same update. Resolves NO if it is equal or lower, or if GPT-6 Astra does not appear in that update. A single instant is evaluated: that update, not any earlier moment. If Artificial Analysis does not publish a Coding Index on or before that date, the market is voided. Source: the Artificial Analysis Coding Index. The market closes on December 31, 2026 at 23:59 Mexico City time (05:59:00 UTC on January 1, 2027).
• When it closesJanuary 1, 2027, 5:59 am UTC
• Verification sourceArtificial Analysis, Leaderboard (https://artificialanalysis.ai/leaderboards/models)