Inkling 让你能控制模型思考深度，975B 参数但只激活 41B，适合做定制化基础模型。
2026年7月15日，Thinking Machines Lab 发布其首个从头训练的模型 Inkling。Inkling 拥有 975B 总参数、41B 激活参数，采用 Mixture-of-Experts 架构，原生支持文本、图像和音频输入，上下文窗口达 1M token。模型权重基于 Apache 2.0 许可开源，实验室明确它并非当下最强模型，而是定位为定制基础，其可控思考努力（Controllable Thinking Effort）是主要差异化功能。
Thinking Machines Lab Releases Inkling: A 975B-Parameter Open-Weights Multimodal MoE With 41B Active Parameters And Controllable Thinking Effort
Thinking Machines Lab released Inkling on July 15, 2026, its first model trained from scratch. The full weights ship under Apache 2.0. It is a 975B-parameter Mixture-of-Experts transformer with 41B active parameters, a 1M-token context window, and native text, image, and audio input. The lab states plainly that Inkling is not the strongest model available, open or closed. It is positioned instead as a customization base, with controllable thinking effort as the practical differentiator. The post Thinking Machines Lab Releases Inkling: A 975B-Parameter Open-Weights Multimodal MoE With 41B Active Parameters And Controllable Thinking Effort appeared first on MarkTechPost .