Pyyan / News / 12 November 2026

Open weightsThinking Machines Lab

Thinking Machines published the weights of Inkling 2 Small, a 594B model with 17B active, under Apache 2.0

500GBaggregate GPU memory the NVFP4 checkpoint needs, such as 2 B300s or 4 H200s

Thinking Machines Lab put the Inkling 2 Small weights on Hugging Face on 12 November under Apache 2.0. The model launched on 5 November as hosted only, on Tinker. It has 594 billion parameters, 17 billion active per token, and a context of up to 1 million tokens. It takes text and images and writes text. It was trained natively in NVFP4, a 4-bit format. The main checkpoint needs at least 500GB of GPU memory. The benchmark scores on its card are the lab's own.

Hugging Facesysnotefrom the source itself

Related

← All the news, newest first