Thinking Machines Lab launches 975B-parameter Inkling AI model

Thinking Machines Lab, founded by former OpenAI executive Mira Murati, released Inkling on July 15, 2026. Inkling is a 975‑billion‑parameter mixture‑of‑experts model with only about 41 billion parameters active per query, a design that balances capability with efficiency. It was trained from scratch on roughly 45 trillion multimodal tokens (text, image, audio, video) and supports a context window of up to one million tokens. The model’s weights are open‑source and can be downloaded via Hugging Face, while the companion Tinker platform allows enterprises to fine‑tune the model on their own data for personalized behavior.

The launch emphasizes customizability over raw size, positioning Inkling as a base for companies rather than a consumer‑focused assistant. Strategic partnerships underpin the effort: Nvidia provides a gigawatt‑scale Vera Rubin compute capacity under a long‑term agreement, and Google Cloud has signed a multibillion‑dollar cloud‑services deal to support the model’s training and deployment.