kimi-k2-turbo-preview Model Output Speed Increased to 60 Tokens per Second
-
On August 22, Moonshot AI announced another major breakthrough in the output speed of its Kimi K2 model. Through the relentless efforts of the engineering team, the kimi-k2-turbo-preview model's output speed has been increased to 60 Tokens per second, with a peak of up to 100 Tokens per second. This significant speed improvement will further optimize user experience and meet high-efficiency demands in more scenarios.
Currently, the kimi-k2-turbo-preview model still enjoys a 50% discount. The price per million Tokens for input (cache hit) is 2.00 RMB, for input (cache miss) is 8.00 RMB, and for output is 32.00 RMB. This promotional price will last until September 1, after which it will revert to the original price.
In the announcement, Moonshot AI expressed gratitude for users' continued support and trust, stating that the company will remain committed to optimizing model performance and further improving the output speed of the Kimi K2 model. Users can visit the Moonshot AI official website (https://platform.moonshot.cn) for more details.
This speed upgrade is not only a technical advancement but also a proactive response by Moonshot AI to user needs. With the continuous improvement in model performance, Kimi K2 will provide users with a more efficient and smoother AI experience, helping them achieve higher productivity and creativity in both work and daily life.