2026•Thinking Machines Lab
Inkling
Open Source
Open Weights
Multimodal AI
Inkling is a multimodal, Mixture-of-Experts transformer model featuring 975B total parameters and 41B active parameters. It is designed as a broad, balanced generalist foundation model capable of reasoning across text, images, audio, and video, with controllable thinking effort for optimized cost and latency. Its multimodal capabilities and efficient, controllable reasoning make it an ideal base for building agentic tools, web applications, and complex, cohesive content artifacts.
975B total / 41B active parameter Mixture-of-Experts architecture
Multimodal reasoning support for text, images, audio, and video
1M token context window for processing large datasets