Thinking Machines Lab releases Inkling 2 Mini, a looped mixture-of-experts model with about 26B total and 3B active parameters and a 1M-token context
On November 20, 2026 Hugging Face announced Inkling 2 Mini from Thinking Machines Lab, a multimodal looped MoE model for local setups with NVFP4 and MXFP8 expert weights, transformers, llama.cpp and MLX support and day-z…