Kimi K2 Thinking, a new reasoning variant of Kimi K2, has recently achieved a significant milestone by surpassing GPT5-codex and Minimax M2 to secure the top position in the Tau2 Bench Telecom agentic benchmark. This achievement marks Kimi K2 Thinking as potentially the new leading open weights model. The model stands out as one of the largest open weights models ever, boasting a total of 1 trillion parameters. This development is a testament to the advancements in AI and the continuous push towards more sophisticated and capable models in the field. The full results are still being processed by AA, but the anticipation for the final scores is high. This news comes from various sources including Twitter and , highlighting the widespread interest and excitement around this achievement. The images and discussions surrounding this release further underscore the significance of this development in the AI community.

标签: none

评论已关闭