Skip to main content

​
What is Fast Agent?

Fast Agent is a Python framework that enables you to create and interact with sophisticated multimodal AI agents and workflows in minutes. It’s the first framework with complete, end-to-end tested MCP (Model Context Protocol) feature support, including Sampling and Elicitations.
Fast Agent is CLI-first, with an optional prompt_toolkit-powered interactive terminal prompt. Responses stream live to the terminal via rich formatting without relying on full-screen curses UIs or external GUI overlays.

​
Key Features

​
MCP Protocol Support

  • Complete MCP Integration: First framework with full MCP feature support including Sampling and Elicitations
  • Transport Diagnostics: Only tool that allows you to inspect Streamable HTTP Transport usage
  • OAuth Support: Built-in OAuth with KeyRing storage for secrets using fast-agent auth command
  • MCP Ping Utility: Optional keep-alive monitoring for client-server connections

​
Multi-Provider Support

Fast Agent provides comprehensive model support across multiple providers:
  • Native Support: Anthropic (Claude), OpenAI (GPT), Google (Gemini)
  • Extended Providers: Azure, Ollama, Deepseek, and dozens more via TensorZero
  • Special Models: Passthrough and Playback LLMs for rapid development and testing

​
Agent Capabilities

Simple Declarative Syntax

Define agents with minimal boilerplate using decorators and simple configuration files

Agent Workflows

Chain, parallel, router, orchestrator, evaluator-optimizer, and MAKER workflows

Multimodal Support

Handle text, images, PDFs, and video with structured outputs

Interactive Development

Chat with agents during and after workflow execution to tune and diagnose

​
Why Fast Agent?

​
Version Control Friendly

Prompts and configurations that define your agent applications are stored in simple files with minimal boilerplate, enabling easy management and version control. 
@fast.agent(
  instruction="Given an object, respond only with an estimate of its size."
)

​
Build Effective Agents

The simple declarative syntax lets you concentrate on composing your prompts and MCP servers to build effective agents, following best practices from Anthropic and other leading AI research.

​
Agent Skills System

Fast Agent supports agent skills (SKILL.md) for reusable agent capabilities that can be shared and composed across your applications.

​
Developer Experience

  • Live Testing: Chat with individual agents and components before, during, and after workflow execution
  • Human-in-the-Loop: Agents can request human input to get additional context for task completion
  • Model Selection: Simple model switching makes testing Model ↔ MCP Server interactions painless
  • Rich Terminal UI: Progress display, tool call visualization, and markdown streaming in the terminal

​
Recent Features

  • Agent Skills (SKILL.md)
  • MCP-UI Support
  • OpenAI Apps SDK (Skybridge)
  • Shell Mode
  • Advanced MCP Transport Diagnostics
  • MCP Elicitations

​
Architecture

Fast Agent follows a component-based architecture:
Fast Agent is designed for rapid prototyping and production deployment. Start with simple agents and scale up to complex multi-agent workflows as your needs grow.

​
Use Cases

  • Automation: Build agents that automate complex tasks using MCP servers
  • Data Analysis: Create agents that analyze data and generate reports
  • Content Generation: Generate social media posts, cover letters, and more
  • Research: Build researcher agents with evaluator-optimizer workflows
  • Development Tools: Create coding assistants with filesystem and tool access

​
Next Steps

Installation

Install Fast Agent and set up your development environment

Quickstart

Build your first agent in under 5 minutes

Build docs developers (and LLMs) love

Get started for freeTalk to us