InjazAI

Thinking Machines Lab released Inkling, an open-weights Mixture-of-Experts model with 975B total parameters and multimodal capabilities, licensed under Apache-2.0.

1 min read

Inkling: Open-weights Model Released by Thinking Machines Lab

Inkling: Open-weights Model Released

FAQ

What is Inkling?

Inkling is an open-weights AI model from Thinking Machines Lab, using a Mixture-of-Experts architecture and supporting multiple modalities.

How does Inkling compare to other models?

Inkling is not a frontier model but competes with open-weights models like NVIDIA Nemotron and Gemma 4, focusing on customizability via Tinker.

Should MENA teams adopt Inkling now?

Yes, teams can start testing Inkling for fine-tuning in multimodal applications, especially with the open Apache-2.0 license.

What are Inkling's key features?

Multimodal capabilities, efficient thinking, availability on Tinker for fine-tuning, and Apache-2.0 license.

Source: Simon Willison (LLM & tools)

AI-assisted content, human-reviewed.