Resolution criteria

This market resolves to YES if, at any point before midnight on December 31, 2027 (anywhere in the world), an AI model developed by a company that doesn't have any model ranked in the top 100 on the LMArena Overall leaderboard at the time of this market's creation (April 24, 2026) appears in the top 50 of the overall leaderboard.

Otherwise, this market resolves to NO.

If there's a tie for Elo score, I will count all models with the same listed score (ignoring the uncertainty and caveats about being "preliminary") as sharing the lowest-numbered rank amongst them. I.e., if the model listed at #51 has the same Elo score as a model listed earlier – which incidentally happens to be the case at the time of market creation – it still counts as "top 50" for the purposes of this market.

A company being renamed, such as from Zhipu to Z.ai, doesn't count.

A stealth model without a listed creator does not count.

The list at the time of market creation is as follows:

1. Anthropic (Rank #1: claude-opus-4-7-thinking)

2. Google (Rank #5: gemini-3.1-pro-preview)

3. Meta (Rank #6: muse-spark)

4. xAI (Rank #8: grok-4.20-beta1)

5. OpenAI (Rank #9: gpt-5.4-high)

6. Z.ai (Zhipu) (Rank #15: glm-5.1)

7. Alibaba (Rank #19: qwen3.5-max-preview)

8. DeepSeek (Rank #20: deepseek-v4-pro)

9. Bytedance (Rank #25: dola-seed-2.0-pro)

10. Moonshot (Rank #26: kimi-k2.6)

11. Baidu (Rank #33: ernie-5.0-0110)

12. Xiaomi (Rank #39: mimo-v2-pro)

13. Meituan (Rank #53: longcat-flash-chat)

14. Amazon (Rank #60: amazon-nova-experimental)

15. Mistral (Rank #82: mistral-large-3)

16. MiniMax (Rank #95: minimax-m2.7)

At the time of market creation, the threshold to be in top 50 is an Elo score of 1439, equaled by

• deepseek-v4-flash-thinking (DeepSeek)

• gemini-3.1-flash-lite-preview (Google)

• gemma-4-26b-a4b (Google)

• gpt-5.1 (OpenAI)

• gpt-5.2 (OpenAI)

The official source for resolution is the Arena leaderboard, available at https://arena.ai/leaderboard/text.

If the leaderboard is discontinued, becomes inaccessible, or changes its methodology in a way that renders "top 50" or "top 100" rankings non-comparable, the market will resolve NA at the discretion of the market creator.

This description was generated by AI.