DataAIHub
DataAIHubNews · Research · Tools · Learning
Back to AI Tools Directory

PocketFlow

Free

100-line minimalist LLM framework for building agents and workflows in pure Python.

Open SourceSelf-hostedPython SDK

Tool Info

Categories
Agents · Frameworks
Developer
The-Pocket
License
Open Source
Official Website
github.com/The-Pocket/PocketFlow
Repository
The-Pocket/PocketFlow
Official Docs

Overview

PocketFlow distills LLM workflow patterns into a tiny Python library.

Developers who understand the code can customize every behavior.

Ideal when LangChain or LangGraph feel over-engineered.

Features

  • Graph-based flows in ~100 lines
  • Node and edge abstractions
  • No heavy dependencies

Pricing

Free tier available
Free (open source)

Pros

  • Extremely lightweight
  • Full control
  • Easy to understand and extend

Best For

Learning agent patternsMinimal production agentsCustom workflow control

When NOT to Use

  • DIY for integrations
  • No enterprise support

Typical Users

Python developersEngineers avoiding heavy frameworks

Related Tools

Google ADK
Agent Development Kit for building multi-agent workflows on Google Cloud and Gemini.
BeeAI
Open framework for building, deploying, and observing production AI agents.
SmolAgents
Minimal agent library from Hugging Face — simple code agents with tool use.

Alternatives

SmolAgents
Minimal agent library from Hugging Face — simple code agents with tool use.
LangGraph
LangGraph is full-featured with persistence and cloud; PocketFlow is a minimal teaching-oriented alternative.

Tags

#agents#minimal#framework

Related Guides

Stay Updated

Get the latest AI news, tools, and engineering guides delivered to your inbox.

Subscribe to Newsletter