GAIA — The Agent Factory.
Build, run, and distribute AI agents that run locally on AI PCs — private by default, no API keys required. Write one in minutes; publish to the Agent Hub and anyone can install it in one click.
Fastest start: download, browse the hub, run an agent — then build your own or embed one in your app. Free · Windows, macOS, Linux · or install via CLI →
Run it, integrate it, or build it.
Pick your level — from a one-click desktop app to building and publishing your own agent. All local, all private.
Run agents
Install the Agent UI desktop app, pick an agent from the Hub, and go — a local AI that sees your files and never calls the cloud.
npm install -g @amd-gaia/agent-uiEmbed agents in your app
Drop a published agent into your own product — a local REST sidecar, an MCP server, or the OpenAI-compatible API. No Python for your users.
npm i @amd-gaia/agent-emailBuild & publish your own
Write an agent with the SDK in a few dozen lines of Python, then publish it to the Hub via a PR so anyone can install it.
gaia agent init my-agentDrop GAIA into your own app.
Integrate whole agents — or individual components — into your third-party app. No Python required for your users; GAIA supplies the building blocks, you ship the experience.
// embed the email agent as a local sidecar
import { fetchBinary, startSidecar } from "@amd-gaia/agent-email";
const { binaryPath } = await fetchBinary({ outDir: "resources" });
const sidecar = await startSidecar({ binaryPath, port: 8131 });There's an agent for that.
Each one is purpose-built for a single job and bulletproof — tested, versioned, and immutable once published.
Prefer the terminal? Start here.
Install the framework and CLI, then build and publish your own agent.
npm install -g @amd-gaia/agent-uiThe desktop app, driven from the CLI.
uv pip install amd-gaiaThe Python SDK + CLI framework.
curl -fsSL https://amd-gaia.ai/install.sh | shOne-line install (Windows: irm …install.ps1 | iex).