> ## Documentation Index > Fetch the complete documentation index at: https://docs.runchat.com/llms.txt > Use this file to discover all available pages before exploring further. # Overview > Build AI workflows visually. Connect generative models, agents, and code on a canvas you can publish as tools or apps. Runchat is a visual canvas for AI workflows. Designers and architects use it to generate concept images, train custom styles, script Rhino, build site research tools, and ship the result as a shareable app. Workflows are built from three nodes: * **Agent**: process data, make decisions, use tools * **Create**: run generative AI models for images, video, and 3D * **Code**: JavaScript, HTML, Rhino Python, Blender Python, or Revit C# Any workflow can be collapsed into a [Tool](/concepts/publishing/libraries) and reused inside other workflows, or given a custom web front-end and published as a public [App](/concepts/publishing/app-view) anyone can use. Works with Rhino, Grasshopper, Blender, SketchUp, and Revit through native [plugins](/plugins/rhino). A Runchat canvas with connected nodes producing rendered images ## Get started [Sign up at runchat.com](https://runchat.com/signup) with Google or email and password. The Runchat Starter plan is free and includes 500 credits to start generating immediately. See [Credits](/concepts/credits) for pricing. Build your first workflow in 10 minutes by generating image prompts in batches ## Learn the fundamentals Worked recipes for rendering, scripting, site research, and more How nodes, data, agents, and publishing fit together Detailed reference for every node and tool Connect your Rhino viewport and geometry to AI workflows ## Watch the tutorials The [Runchat 101: Foundations](https://www.youtube.com/playlist?list=PLgT771y9VleCblM25Swp-LM-xzPiFDbab) playlist on YouTube walks through the platform end to end, from the editor and nodes through to publishing.