Will a Chinese company have the best AI model by December 31?
As of July 14, 2026, the market gives “Will a Chinese company have the best AI model by December 31?” a 91% chance of NO.
| YES odds | 10% |
|---|---|
| NO odds | 91% |
| Volume | $52,622 |
| Closes | December 31, 2026 |
Updated July 14, 2026 · Live data from Polymarket
About this market
US labs maintain clear leads on frontier benchmarks as of mid-2026, with Anthropic’s Claude Opus 4.8 and OpenAI’s GPT-5.5 family topping composite scores and reasoning evaluations while Chinese models from DeepSeek, Alibaba’s Qwen series, and Moonshot’s Kimi trail by small but consistent margins. Strong US advantages in proprietary training compute, talent concentration, and rapid iteration cycles underpin trader consensus around the 93.5% “No” probability. Chinese developers have narrowed the gap through cost-efficient open-weight releases and high token usage, yet consistent outperformance on the hardest closed-model leaderboards remains rare. A credible challenge would require a verified Chinese model to surpass current US leaders on multiple public benchmarks before year-end, an outcome viewed as improbable given current trajectories.