Thinking Machines Lab launches Inkling, 975‑billion‑parameter open‑weight AI model

Thinking Machines Lab, the AI startup founded by former OpenAI chief technology officer Mira Murati, released its first model, Inkling, on 15 July 2026. Inkling is a multimodal Mixture‑of‑Experts transformer with 975 billion total parameters, of which about 41 billion are active for any given task. It was trained from scratch on 45 trillion tokens of text, images, audio and video and supports a context window of up to one million tokens.

The model is released under an Apache 2.0 license with its full weights publicly available on Hugging Face. Revenue is generated not from API calls but from the company’s Tinker platform, which lets developers fine‑tune Inkling for specific business needs. The startup highlighted a Bridgewater Associates collaboration that used Tinker to fine‑tune a financial‑specialised version of the model at a fraction of the cost of proprietary alternatives.

Thinking Machines acknowledges that Inkling “is not the strongest overall model available today,” but it aims to provide a balanced, customizable base that rivals leading Chinese open‑weight models while reducing compute cost. The launch represents a strategic push to offer a Western alternative in the open‑weight AI ecosystem and could shift enterprises toward locally‑controlled, fine‑tuned AI solutions.