Resolution criteria

• This market consists of multiple independent options representing different score thresholds for OpenAI's "Astra" model on the Artificial Analysis Intelligence Index.

• Source of Truth: The official model scores listed on the Artificial Analysis LLM Leaderboard.

• Evaluation:

  • If multiple configurations or effort levels of the OpenAI Astra model are listed (e.g., "Astra (xhigh effort)" vs. "Astra (low effort)"), the highest listed Intelligence Index score among them will be used for resolution.

  • If the model has not been graded on the Artificial Analysis Intelligence Index by the closing date, the market end date will be extended until December 31st.

Background

• OpenAI's next-generation model family, code-named Astra (frequently discussed as GPT-6 Astra), is designed for long-horizon planning, advanced mathematics, and autonomous agent capabilities.

• The Artificial Analysis Intelligence Index is an aggregated, independent benchmark score designed to evaluate frontier model capabilities.

• Current model scores on the Artificial Analysis Intelligence Index for context:

  • Claude Fable 5.1: 66

  • Claude Opus 5: 63

  • GPT-5.6 Sol: 61