What We Do

ClawVision develops open models and tools for spatial understanding. Our work spans computer vision, multimodal AI, and embodied intelligence — from training vision-language models to deploying perception systems in real-world environments.

We believe the best research happens in the open. We publish our findings, release our models, and share the lessons learned along the way.

Research Philosophy

  • Open by default — Models, code, and data released publicly whenever possible.
  • Practical impact — We care about systems that work in the real world, not just on benchmarks.
  • Reproducibility — Every claim backed by code you can run.
  • Community-driven — Built with and for the broader AI research community.

Contact

Interested in collaborating or have questions about our work? Reach out at research@clawvision.org or find us on GitHub.