Will a Chinese company have a top ___ AI model by December 31?
The event asks whether the AI model with the single highest score on the LMArena (Chatbot Arena) text leaderboard will be owned by a Chinese company. The score used is the "Score" column under "Text Arena | Overall" with style control off. A tie for first place does not count, and the wording "at any point between market creation and December 31" leaves it unclear whether one qualifying check is enough.
Leading right now: #3 6%, #1 2%.
| Market | Chance of yes | Deadline |
|---|---|---|
| Will a Chinese company have one of the top 3 AI models by December 31? | 5.5% | Jan 1, 2027, 4:59 UTC |
| Will a Chinese company have the best AI model by December 31? | 2.3% | Jan 1, 2027, 4:59 UTC |
Odds read from Polymarket at 21:54 UTC on Nov 21, 2026. $335K traded so far, $2K in the last 24 hours.
How this market resolves
Yes: A Chinese company owns the model with the highest score in the "Score" column of the Text Arena Overall leaderboard (style control off) at the check time, and that score is strictly higher than every other model's score.
No: A non-Chinese company's model has the highest score at the check time, or a Chinese model only ties for first place. Companies not headquartered in China, or not primarily owned or operating within the Chinese technology ecosystem, are not treated as Chinese.
When it really ends
December 31, 2026, 11:59 PM ET. New York is on standard time (UTC-5) on that date, so this is 2027-01-01 04:59 UTC. The endDate field in the data (2027-01-01T04:59:00Z) matches the rules. The rules also say the check can happen "at any point between market creation and" that deadline.
What counts
- The "Score" column on https://lmarena.ai/leaderboard/text under "Text Arena | Overall" with style control off.
- Models owned by the named Chinese companies: Alibaba, ByteDance, Baidu, Moonshot, Z.ai, DeepSeek, Meituan, Xiaomi, StepFun, Tencent and MiniMax.
- Other companies that are primarily Chinese, since the list says "include but are not limited to".
- A Chinese model whose score is strictly higher than every other model's score at the check time.
What does not count
- Ties: a Chinese model level with another model at the top does not qualify.
- Rankings on other leaderboard tabs or categories, such as vision, coding or other Arena boards.
- Scores with style control turned on.
- Companies not headquartered in China or not primarily owned or operating within the Chinese technology ecosystem.
Fine print people miss
- The rules text provided describes only the highest-scoring model. The top-3 market's own rules are not shown, so how it counts a top-3 place is not stated here. Read that market's rules before assuming it follows the same test.
- The wording "at any point between market creation and December 31" is ambiguous. It may mean one qualifying check is enough, or it may mean the final check decides. The rules do not say who performs the check or when.
- If the leaderboard is unavailable at check time, the market stays open and resolves on the first check after it returns. If it is permanently unavailable, another resolution source is used.
- The rules do not say how ownership is decided when a model is jointly developed or backed by several companies. Only the "primarily Chinese" test is given.
What moves the price
- A new flagship model release from a major Chinese lab that scores near the top would raise the chance of Yes.
- A new release from a leading non-Chinese lab would raise the chance of No, because it could push the top score further away.
- Early scores for a new model can shift as more votes arrive, so a model rising or falling on the board changes the outlook.
- Changes to leaderboard methodology, default views or the models listed could move the top spot either way.
- Leaderboard downtime or a ruling on an unclear ownership case could affect the timing or the outcome.
Put a bot on this market
A PROWL bot reads the news around the clock, judges each post against these exact rules, and trades inside the limits you set. It starts in paper mode.
Questions people ask
What leaderboard decides the Chinese AI model market?
The LMArena (Chatbot Arena) text leaderboard at lmarena.ai/leaderboard/text. The market uses the "Score" column under "Text Arena | Overall" with style control off.
When does the market end?
At 11:59 PM ET on December 31, 2026, which is 04:59 UTC on January 1, 2027. The rules also mention a check at any point between market creation and that time.
Which companies count as Chinese?
The rules name Alibaba, ByteDance, Baidu, Moonshot, Z.ai, DeepSeek, Meituan, Xiaomi, StepFun, Tencent and MiniMax. The list is not exhaustive. Companies not headquartered in China or not primarily Chinese do not count.
What happens if a Chinese model ties for first place?
It does not count as Yes. The rules require a strictly higher score than all other models, so ties do not qualify.
Does style control matter?
Yes. Only scores with style control off are used. Scores shown with style control on are not the resolution figure.
What if the leaderboard website is down?
The market stays open until the site returns and resolves on the first check after that. If the site is gone for good, another resolution source is used.
More AI markets explained
Written from the market's own rules text on Nov 16, 2026. If the rules on Polymarket differ from this page, the rules on Polymarket govern.