DeepSeek's commanding position stems from the mid-October release of its V4.1 Flash 1017 large language model on OpenRouter, which quickly captured the top model ranking with tens of trillions of tokens processed weekly through superior cost efficiency, a 1M-token context window, and gains in agentic and coding benchmarks. Low per-token pricing combined with open-weight MIT licensing drove heavy developer routing for this and prior Flash variants, pushing the lab's tracked share to roughly 24-32%—well ahead of Z.ai, MiniMax, and U.S. labs. Sustained momentum from iterative efficiency improvements has reinforced trader consensus. A rival model launch or unexpected capacity constraints could still alter the week-of-October-19 outcome before resolution.
基于Polymarket数据的AI实验性摘要。这不是交易建议,也不影响该市场的结算方式。 · 更新于1H6H1D7D30DALL
View resolved













警惕外部链接哦。
警惕外部链接哦。
常见问题