975B MoE open-weights model with 41B active parameters
Inkling open-weights model now on Databricks platform
Thinking Machines Lab released Inkling, a new multimodal mixture-of-experts model that accepts text, image, and audio inputs and produces text outputs. It features controllable reasoning effort, a 975B total parameters with 40B active per token, 1M token context window, and architectural innovations including query-conditioned relative attention, short causal convolutions, and a shared expert sink. Together AI offers access through serverless and OpenAI-compatible APIs.
Thinking Machines Lab released Inkling, a large open multimodal model with ~1T parameters and 1M context window, natively accepting image, text, and audio inputs. It also released Inkling-Small with 276B total and 12B active parameters, and Inkling-Small-NVFP4 variant. The models are available on Hugging Face with day-0 support in transformers, SGLang, vLLM, and llama.cpp.
