Kimi 刚发了 K3，能把图片视频转成可交互 3D，还能边写代码边看效果，挺实用的。
Kimi 发布了 K3 模型，结合 3D 推理、编程和视觉理解能力。它能够将文本概念、图片或视频转化为可玩的交互式 3D 体验。K3 实现了"vision in the loop"机制，模型可在代码和实时截图之间来回迭代。
Kimi K3 结合了 3D 推理、编程和视觉能力。 它可以把概念、图片和视频转化为可玩的交互体验。 Kimi 表示，K3 实现了真正的 “vision in the loop”： 模型可以在代码和实时截图之间来回迭代。 Your browser does not support the video tag. 🔗 View on Twitter 💬 1 🔄 0 ❤️ 0 👀 36 📊 1 ⚡
This workflow demonstrates a practical application of combining multiple AI models to automate content creation, potentially saving hours of manual work daily. For content creators and AI enthusiasts in ASEAN, this approach offers a template for building similar automated news systems using locally accessible models like GLM 5.3, which could be particularly valuable for regional language content creation.
For users of the Kimi Code platform, the K2.8 Preview model offers improved performance and a larger context window, making it a better option than the previous K2.7 Code version.
看热闹的来了，K3 刚出就有人爆 K4 料，参数量可能冲到 5-6T，还牵扯到 Blackwell 算力抢购，想了解下一代卷王的可以看看。
WorkBuddy把AI订阅安排得明明白白，免费的和高性价比的一起用，每天5000万token不心疼。
