amd/gaia

Build AI agents for your PC

What it solves

GAIA provides a framework for building and deploying intelligent AI agents that run locally on AMD Ryzen AI hardware. This eliminates reliance on cloud inference, reducing costs and improving data privacy for sensitive workloads (such as healthcare or finance) and enabling air-gapped deployments.

How it works

GAIA offers a base agent class that handles tool orchestration, state management, and error recovery. It allows developers to create agents by defining system prompts and registering custom tools. The framework is optimized for AMD Ryzen AI processors, utilizing the NPU and iGPU for hardware-accelerated inference. It also includes a C++17 port for native application embedding and a desktop UI for chat, document indexing, and tool execution.

Who it’s for

Developers building AI agents who want to leverage local hardware acceleration on AMD Ryzen AI systems, as well as enterprises requiring high privacy and security for their AI deployments.

Highlights

  • Local-First Inference: Hardware-accelerated performance using NPU and iGPU on AMD Ryzen AI processors.
  • RAG System: Built-in document indexing and semantic search supporting over 50 file formats.
  • Multimodal Capabilities: Integrated voice interaction (Whisper ASR and Kokoro TTS) and vision models (Qwen3-VL-4B) for image text extraction.
  • Extensibility: Supports MCP (Model Context Protocol) integration for external tools and a plugin system for distributing agents via PyPI.
  • Cross-Language Support: Available in both Python and C++17.

Related

  • Project
    erthorpabar/pabar_agent_runtime

    An AI agent runtime framework that integrates chat, RAG, and agentic loops to enable professional automation tasks across desktop and mobile platforms.

  • Project
    MaestroError/LarAgent

    An AI agent development framework for Laravel that allows developers to create and maintain AI agents using an Eloquent-like API and Laravel-style tooling.

  • Project
    mudler/LocalAGI

    A self-hostable, no-code AI agent platform that allows users to create private, local AI automations and agents on their own hardware.

  • Project
    BigSweetPotatoStudio/HyperChat

    A local AI Agent platform that implements an "AI as Code" philosophy, allowing project-specific AI assistants to be configured, versioned, and managed locally via files.

  • Project
    i-am-bee/beeai-framework

    A comprehensive toolkit for building production-ready autonomous agents and multi-agent systems in Python and TypeScript, featuring integrated RAG, memory, and multi-LLM support.