Our mission is to build AI that extends human will and judgment. Today we advance that mission by releasing Inkling, a model trained from scratch with full weights available for customization. Inkling is a Mixture-of-Experts transformer with 975B total parameters and 41B active, supporting a 1M token context window. It was pretrained on 45 trillion tokens of text, images, audio, and video. Alongside it we share Inkling-Small, a lighter model with 12B active parameters for lower cost and latency.

Inkling reasons natively over text, images, and audio with controllable thinking effort, balancing performance with token efficiency. It reaches a given score at fewer tokens — matching Nemotron 3 Ultra on Terminal Bench 2.1 at roughly a third of the tokens. On Design Arena's Agentic Web Dev leaderboard, Inkling ranks among the strongest open-weight models.

Inkling is not the strongest overall model available, but its combination of multimodal capabilities, efficient thinking, and availability on Tinker for fine-tuning makes it a good open-weights base. We're adding the Inkling Playground in the Tinker console for developers to chat with the model. Inkling is just the start: our first release in a model family we will continue to build on.