BigHugger
GH Repository · FareedKhan-dev

all-agentic-architectures

35 production-grade agentic AI architectures (Reflexion, LATS, GraphRAG, MemGPT, Voyager, BrowserAgent, ...) — a Python library and runnable textbook with multi-provider LLM support and a 17-task benchmark leaderboard.

stars
4,516
30-day movement
+11037/day
Related entries
60
Connections
2
llmpythonlangsmithlanggraphagent-applangchainJupyter Notebookai-agentsagentic-ai

A Python library and runnable textbook presenting 35 agentic AI architectures, including Reflexion, LATS, GraphRAG, MemGPT, Voyager, and BrowserAgent. It ships with multi-provider LLM support and a 17-task benchmark leaderboard.

You want to study and run reference implementations of many agent architectures from one repo instead of reimplementing each from papers.

Use it to

  • Run and compare 35 agent architectures in one environment
  • Benchmark agents across the included 17-task leaderboard
  • Study implementations like Reflexion, LATS, GraphRAG, and MemGPT
  • Swap between LLM providers while prototyping agent designs
  • Launch a devcontainer to reproduce the notebook-based textbook

For Developers and researchers learning or benchmarking LLM agent architectures

Role
agent-app
Language
Jupyter Notebook
Licence
MIT
Forks
769
Open issues
4
Last push
2026-06-22
Latest release
v0.3.0 · 2026-05-28
topicsagentic-aiai-agentslangchainlanggraphllmbenchmarking