AI AGENT FRAMEWORKS
FEATUREDORCHESTRATIONOPEN SOURCE SELF-HOSTABLE
LangGraph
by LangChain
Stateful, controllable agent orchestration as graphs.
94HYLAND SCORE
9.8k views this week
Overview
LangGraph models agents as state machines: cycles, branches, persistence, and human-in-the-loop are first-class. It is the most widely adopted framework for production agents that need reliability, retries, and long-running state.
Key features
- Graph-based agent state machines
- Built-in persistence and checkpointing
- Human-in-the-loop interrupts
- LangSmith tracing and evals
- LangGraph Platform for deployment
Tags
PythonTypeScriptStatefulHuman-in-the-loop
Related in AI Agent Frameworks
VERIFIED
Semantic Kernel
Microsoft
84HYLAND
Enterprise AI orchestration for .NET and Python.
OrchestrationOpen Source Self-host
5.2k/wk
VERIFIED
Haystack Agents
deepset
78HYLAND
Production pipelines with agentic loops.
OrchestrationFreemium Self-host
3.1k/wk