Inkling brings Thinking Machines into open multimodal models

Hugging Face published an introduction to Inkling from Thinking Machines Lab. The model natively accepts text, image, and audio inputs, supports a large context window, and includes agentic capabilities. The publication describes deployment formats and support across major inference libraries.

Context

A multimodal release is useful when the tooling can actually run it within a team’s hardware budget. Parameter totals alone say little about that experience. Input types, quantization, runtime support, and the quality of cross-modal reasoning all belong in the deployment decision.