模型
也叫：kimi-k2.5
Fireworks平台支持的先进语言模型
最近变化
2026-05-20 · 基线模型在5次调用中约1次输出畸形，导致多步骤工作流重试
Kimi K2.5 被反复提及时，通常意味着它正在影响产品路线、开发者工作流或 AI 产业判断。这个页面把分散材料合并成一个可持续更新的观察入口。
We ran 720 browser agent tasks with @nottecore across frontier models. One baseline model produced...
Fireworks AI(@FireworksAI_HQ) · 8.5 分
The @cursor_ai team shipped Composer 2 and now Composer 2.5 on the same Kimi K2.5 base model. Perfor...
Fireworks AI(@FireworksAI_HQ) · 7.2 分
Composer 2.5 is built on the same open-source base as Composer 2, Moonshot’s Kimi K2.5.
Cursor(@cursor_ai) · 2.5 分
已收录 3 篇与「Kimi K2.5」相关的 AI 资讯和分析。
Fireworks AI tests show baseline models had 20% retry rates in browser agent tasks, while Kimi K2.5/GLM-5/MiniMax M2.5 achieved near-zero retries with stable latency, directly impacting production system costs/delays/reliability.
入选理由：基线模型在5次调用中约1次输出畸形，导致多步骤工作流重试
Cursor AI launched Composer 2.5 on the Kimi K2.5 base model, achieving 85% performance gains from reinforcement learning, with Fireworks AI providing the RL infrastructure for scalable deployment.
入选理由：Composer 2.5基于Kimi K2.5模型，性能显著提升，85%的算力增益来自强化学习（RL）。
Cursor's official Twitter account posted a brief statement announcing Composer 2.5 is built on the same open-source base as Composer 2, referencing Moonshot's Kimi K2.5, but provided no technical details or architectural explanations.
入选理由：Composer 2.5 与 Composer 2 共享同一开源代码基础
与「Kimi K2.5」经常一起出现的 AI 术语。
💡 想追踪「Kimi K2.5」的长期趋势？去 实体雷达 · Kimi K2.5 查看详细分析和跨材料问答。
