Tnkr is a collaborative platform for people building and exploring open-source robotics projects. It brings together the four components of physical intelligence—hardware, software, data and models—so developers and makers can document their work, share it with others and improve projects through community contributions. Builders can create step-by-step assembly instructions, manage bills of materials, document code and sensor integrations, and connect with tools such as Onshape, SolidWorks and GitHub. The platform also supports collecting real-world operational data and feeding it into model-training workflows, as well as exploring and deploying models and policies for specific hardware.

Tnkr offers interactive project exploration, including 3D visualizations, and helps contributors rebuild and remix projects ranging from humanoid robots to quadrupeds. Its AI assistant, Leonardo, analyzes first-person build videos, CAD and code to generate documentation, suggest improvements and help troubleshoot. Leonardo can also turn assembly videos into guided instructions with AI assistance. By combining project knowledge, documentation and collaboration tools, Tnkr aims to make robotics builds easier to reproduce and develop.