Loading page…
Productivity / product dossier
A mixture-of-experts model for coding, visual tasks, and agents, available for inference and fine-tuning.
Product brief
Inkling 2 Small is Thinking Machines Lab's new 594B-parameter MoE model with 17B active, built for long-horizon coding, visual tasks, and agents. Available on Tinker for inference and fine-tuning.
Why we selected it
A substantive model release for builders and researchers: a 594B-parameter MoE with 17B active, targeting long-horizon coding, visual tasks, and agents. Availability for both inference and fine-tuning on Tinker makes it4
Product preview saved with our daily selection
Capability scan
Only capabilities supported by the product information we collected are listed here.
Built for coding tasks that extend over a long horizon.
Targets visual tasks, with the stated ability to see and refine what it builds.
Designed for use in agents alongside its coding and visual task capabilities.
Available on Tinker for both inference and fine-tuning.
Best-fit use cases
FAQ
Inkling 2 Small is a model from Thinking Machines Lab built for long-horizon coding, visual tasks, and agents.
It is a mixture-of-experts (MoE) model with 594 billion total parameters and 17 billion active parameters.
Inkling 2 Small is available on Tinker for both inference and fine-tuning.