Cheshire Cat

Cheshire Cat is an artificial intelligence agent framework that lets developers build and extend agents using Python.

Cheshire Cat is an artificial intelligence agent framework designed for learners and creatives. It lets users build custom agents, tools, directives, and hooks using pure Python. Developers can run the framework locally, connect various large language models, and extend functionality with plugins stored in simple folders. The system supports multi-agent chats, native Model Context Protocol servers, and live streaming over the AG-UI protocol. It also includes a built-in web user interface and an extensible application programming interface with role-based authentication.

Cheshire Cat at a glance

Best for: Software engineer
Runs on: self-hosted, API, Web
Pricing: Free
Released: February 2023
Category: AI agent tool
Website: cheshirecat.ai

Key features

Python-based agent framework
Multi-agent chat support
Native Model Context Protocol integration
Live streaming over AG-UI
Model swapping for OpenAI, Anthropic, Gemini, and others
Plugin system using simple folders
Built-in web user interface
Extensible application programming interface with authentication

Use it for

Build custom artificial intelligence agents
Connect Model Context Protocol servers
Run multiple agents in a single chat
Develop plugins to add tools and endpoints

Alternatives to Cheshire Cat

Atomic Agents — Build predictable, modular AI applications with consistent, reliable outputs.
Webwright — Enables code-based web automation with reusable and rerunnable scripts.
MyAgent — It features built-in permission controls, long-term memory, and skill self-evolution.
Goose — Runs locally on your machine with support for 70+ extensions and any LLM.
Beeuni — Agents retain context and improve skills over time through a controlled review process.
AGiXT — Builds autonomous agents with integrated memory and web capabilities.
Kungfu — Maintains work state and context when switching between different AI agents.
Rakazo — Complete ownership of data, models, and execution environment.
Bindu — Builds interoperable AI agents with A2A Protocol compliance and payment integration.
labs OO Agents — Combines prompts, tools, and workflows into a single Python class.
II-Agent — It provides a fully open-source, transparent, and auditable agent foundation.
AG2 — Builds and orchestrates multi-agent AI systems with universal interoperability.

All Cheshire Cat alternatives