Resolution criteria

• Source of Truth: This market will resolve based on the official "Text Arena - Overall" leaderboard standings of the LMSYS Chatbot Arena, hosted at Hugging Face Chatbot Arena or Arena AI.

• Timing: The standings will be evaluated based on the leaderboard state on October 1, 2026. If the live site is down, the most recent update or snapshot around that date will be used.

• Winning Lab: Resolution is determined by the organization/developer of the model ranked #1 overall.

• Ties: In the event of an exact tie in Elo score at the #1 spot, the model listed first (occupying the top row) in the default leaderboard view will determine the winning lab.

• Answer Mapping:

  • Anthropic: Resolves YES if the #1 model is developed by Anthropic.

  • OpenAI: Resolves YES if the #1 model is developed by OpenAI.

  • Google: Resolves YES if the #1 model is developed by Google (or Google DeepMind).

  • SpaceXAI: Resolves YES if the #1 model is developed by xAI (the AI lab founded by Elon Musk, creators of Grok).

  • Others: Resolves YES if any other entity (e.g., Meta, Moonshot, Alibaba, Mistral, Xiaomi, etc.) holds the #1 spot.

Background

• Chatbot Arena utilizes crowdsourced blind A/B battles to rank large language models using an Elo-style rating system.

• As of September 2026, Anthropic's Claude Fable/Opus series, OpenAI's GPT-6 models, Google's Gemini, and xAI's Grok are frequently updated, causing rapid shifts at the top of the overall leaderboard. This market tracks which of these frontier labs will lead the leaderboard at the start of October.