ToolAgents
ToolAgents is a lightweight and flexible framework for creating function-calling agents with various language models and APIs. It provides a unified interface for integrating different LLM providers and executing function calls seamlessly.
Overview
ToolAgents aims to provide a consistent way to interact with different language model providers while maintaining a clean, easy-to-use interface for building AI agents that can use tools and functions. Whether you're using OpenAI, Anthropic, Mistral, or other providers, ToolAgents offers a unified approach.
graph TB
User[User] --> |Sends request| Agent
subgraph "ToolAgents Framework"
Agent[Chat Tool Agent] --> |Formats messages| Provider
Agent --> |Registers tools| Tools
Provider[Provider API] --> |Calls| LLM
LLM --> |Response with tool calls| Agent
Agent --> |Executes| Tools
Tools --> |Returns results| Agent
end
Agent --> |Final response| User
class User,LLM external
class Agent,Provider,Tools internal
Key Features
- Multiple LLM Provider Support: Easily switch between OpenAI, Anthropic, Mistral, and other providers
- Unified Message Format: Maintain your chat history across different LLM providers
- Flexible Tool Creation: Build tools using Pydantic models, functions, or OpenAI-style function specs
- Streaming Support: Get streaming responses for a more responsive experience
- Memory Management: Manage chat history and implement memory features
- Advanced Agent Capabilities: Create multi-agent systems, research agents, and more
Quick Links
- Installation Guide - Get started with ToolAgents
- Basic Usage - Learn the core concepts
- Custom Tools - Create your own tools
- Examples - Explore practical applications