> ## Documentation Index > Fetch the complete documentation index at: https://docs.iblueprint.ai/llms.txt > Use this file to discover all available pages before exploring further. # What is iBlueprint? Blueprints, nodes, and AI chains > iBlueprint is a visual AI workflow platform where you build, run, and share Blueprints — sequential chains of AI, API, and logic nodes. iBlueprint is a visual AI workflow platform that lets you chain together AI models, external APIs, custom code, and conditional logic into automated pipelines called **Blueprints**. You build Blueprints in a drag-and-drop editor without writing infrastructure code, then trigger them manually, on a schedule, via webhook, or from the CLI — and iBlueprint handles the execution, variable passing, and result tracking. ## What is a Blueprint? A Blueprint is a self-contained, runnable workflow made up of **nodes** connected in a linear chain. Each node receives input from the node before it, processes that input, and passes its output to the next node in the sequence. You define the chain once and can re-run it with different variables at any time. ## Node types iBlueprint supports five node types that you can mix and match in any Blueprint: | Node type | What it does | | --------------- | ------------------------------------------------------------------------------------------------------- | | **Prompt** | Sends a prompt to an AI model (OpenAI, Anthropic, Gemini, or Ollama) and returns the generated response | | **API** | Makes an HTTP request to any external API with dynamic parameters drawn from earlier nodes | | **Function** | Executes custom JavaScript or TypeScript code for data transformation or business logic | | **MCP** | Connects to a Model Context Protocol server to retrieve contextual information for downstream nodes | | **Conditional** | Branches execution based on a condition, letting you build decision trees inside a single Blueprint | ## AI provider support iBlueprint connects to the four major AI provider families out of the box. You bring your own API keys and store them securely in **Settings → AI Connections**. Supported providers include: * **OpenAI** — GPT-4o, GPT-4, GPT-3.5, and other OpenAI models * **Anthropic** — Claude 3 and later model families * **Google** — Gemini 1.5 Pro, Gemini Flash, and related models * **Ollama** — Self-hosted open-source models running on your own infrastructure You can connect keys for multiple providers and set a default per provider. Blueprint prompt nodes let you choose the provider and model independently for each step in your chain. ## Where you can use iBlueprint iBlueprint is available across platforms so you can build, manage, and run Blueprints wherever you work: * **Web app** — The full visual editor at [app.iblueprint.ai](https://app.iblueprint.ai) * **CLI** — The `iblueprint` command-line tool for scripting and CI/CD pipelines * **Mobile app** — A React Native companion for monitoring executions on the go * **API** — A tRPC-based API at `/api/trpc` for programmatic access Create your first Blueprint and run it in under five minutes. Learn how to build visual AI chains and work with all five node types. Connect OpenAI, Anthropic, Gemini, Ollama, and MCP servers. Call iBlueprint programmatically to list, create, and execute Blueprints.