Documentation — FastAIAgent
Documentation

Build, debug, evaluate, and operate agents.

The only SDK with fork-and-rerun Agent Replay. Start with a five-minute agent, then follow the loop into tracing, evaluation, and operations.

Get started → API reference
$ pip install fastaiagent

from fastaiagent import Agent, LLMClient

llm = LLMClient(provider="openai", model="gpt-4o")
agent = Agent(name="my-agent", system_prompt="You are a helpful assistant.", llm=llm)

result = agent.run("What is the capital of France?")
print(result.output)
print(result.trace_id)  # every run is traced — use this ID for replay

$ pip install 'fastaiagent[ui]' && fastaiagent ui   # → http://127.0.0.1:7842
{{ s.n }}↗

{{ s.title }}

{{ s.body }}

{{ t }}
Tutorials
Build Your First Agent in 5 Minutes↗ Trace a LangChain Agent in 2 Minutes↗ Debug with Agent Replay↗ Memory, End to End↗
Templates
Deep Research Agent↗ Financial Extraction (worked example)↗ Regression from Trace↗
Migration guides
From LangGraph↗ From LangSmith↗ From Langfuse↗ From MLflow↗