Kimi K3.1 Pro在Artificial Analysis Intelligence Index评测中获得55分，位列同价位模型平均水平之上1。这一成绩明显高于同价位模型的中位数27分1。该模型由Kimi公司于2026年10月23日发布1，配备2.8万亿参数规模1，支持文本和图像输入1，具有100万token的上下文窗口1。
在性能表现方面，Kimi K3.1 Pro的首token时间为3.20秒，优于同类平均3.62秒1；但输出速度为73.5 token/秒，低于同类中位数81.1 token/秒1。定价方面，该模型输入成本为每百万token $2.00，输出成本为每百万token $10.001。在Intelligence Index评测中，该模型生成了150万token的输出，高于同类中位数87万token1。
Kimi K3.1 Pro has achieved a score of 55 on the Artificial Analysis Intelligence Index, significantly outperforming models in its price category, where the median score stands at 27 points 1. Released by Kimi on October 23, 2026, the model supports multimodal input including text and images, and features a context window of 1.0 million tokens 1.
The model is priced at $2.00 per million input tokens and $10.00 per million output tokens, with first-token latency of 3.20 seconds, which exceeds the category average of 3.62 seconds 1. However, its output throughput of 73.5 tokens per second falls below the median performance of 81.1 tokens per second for comparable models 1. Kimi K3.1 Pro operates with 2.8 trillion parameters and employs chain-of-thought reasoning for inference tasks 1. During the Intelligence Index evaluation, the model generated 150 million output tokens, surpassing the median output of 87 million tokens across its peer group 1.
评论
还没有评论，欢迎留下第一条。