Thinking Machines Launches Inkling, Open-Weight Multimodal AI Model.

@TheRundownAI· July 15, 2026 View original

Summary

Thinking Machines has released Inkling, an open-weight, multimodal AI model featuring a 1M-token context window and native reasoning across text, images, and audio. The model's full weights are available on Hugging Face, with fine-tuning supported through Tinker, positioning it as a customizable base model.

Thinking Machines has introduced its inaugural AI model, named Inkling, which is now available as an open-weight, multimodal system. This new model boasts impressive features, including a context window capable of handling up to one million tokens and the ability to perform native reasoning across various data types such as text, images, and audio. A notable design choice is its "controllable thinking effort," allowing users to manage computational resources. The company has made the complete model weights accessible on Hugging Face, facilitating broad adoption and experimentation. Furthermore, developers can fine-tune Inkling using the Tinker platform. While Thinking Machines acknowledges that Inkling may not be the most powerful model across all benchmarks, its design prioritizes serving as a robust and adaptable foundation for specialized applications and custom development.

Why it matters

The launch of an open-weight, multimodal model with a large context window and customization options provides developers and enterprises with a flexible foundation for building specialized AI applications. This can accelerate innovation and reduce reliance on proprietary systems.

How to implement this in your domain

  1. 1Download Inkling's weights from Hugging Face to begin local experimentation and integration.
  2. 2Explore Tinker for fine-tuning Inkling with proprietary datasets to create domain-specific models.
  3. 3Develop multimodal applications that leverage Inkling's text, image, and audio reasoning capabilities.
  4. 4Evaluate Inkling's performance against existing models for specific use cases, focusing on its customization potential.

Who benefits

Software DevelopmentMedia & EntertainmentHealthcareRetailEducation

Key takeaways

  • Inkling is an open-weight, multimodal AI model from Thinking Machines.
  • It features a 1M-token context window and native reasoning across text, images, and audio.
  • The model is designed for customization and fine-tuning.
  • Its availability on Hugging Face and Tinker promotes broad developer access.

Original post by @TheRundownAI

"Thinking Machines just launched their first model, an open-weight, multimodal system called Inkling. - Up to a 1M-token context window - Native reasoning across text, images, and audio - Controllable thinking effort - Full weights on Hugging Face, with fine-tuning through Tinker…"

View on X
Thinking Machines Launches Inkling, Open-Weight Multimodal AI Model.

Originally posted by @TheRundownAI on X · view source

Want to go deeper?

Turn these trends into skills with Learnijoy's hands-on AI & tech courses.

Explore courses