Skip to main content
Hypertic is a straightforward, easy, and fast way to build AI agents. It’s a lightweight Python framework that lets you create agents with any large language model, tools, and knowledge bases, while maintaining context and handling complex tasks. Here are some of the key features of Hypertic:

Tools & MCP

Create custom tools with Python functions or connect to MCP servers.

Memory & Context

Store conversation history and maintain context between interactions.

Model Agnostic

Switch between OpenAI, Anthropic, Google, and others without changing your code.

Knowledge Base

Build an agent that can search and use your documents and knowledge bases.

​
Why Hypertic?

Straightforward: Create an agent in minutes with a simple, intuitive interface. No complex abstractions or layers to learn, just pass your model, tools, and memory to the Agent class and you’re ready to go. Easy: Use any LLM provider (OpenAI, Anthropic, Google, etc.) with the same interface. Add tools with a simple @tool decorator. Configure memory with a single parameter. Everything follows Python conventions you already know. Fast: Get started immediately with minimal setup. The lightweight framework has minimal dependencies and overhead, so your agents start quickly and run efficiently without unnecessary bloat. Lightweight: Built with a focus on simplicity and performance. No heavy dependencies or complex abstractions, just what you need to build production-ready AI agents.

​
Example

Here’s an example of an Agent that uses a tool and manages conversation history: 
from hypertic.agents import Agent
from hypertic.tools import tool
from hypertic.models import OpenAI

# Define a custom tool
@tool
def get_weather(city: str) -> str:
    """Get weather for a city."""
    return f"Sunny, 72°F in {city}"

# Create agent with tools
model = OpenAI(model="gpt-4")
agent = Agent(
    model=model,
    instructions="You are a weather assistant.",
    tools=[get_weather],
)

# Use it
result = agent.run("What's the weather in San Francisco?")
print(result.content)
If you’re new to Hypertic, follow the quickstart to build your first Agent or explore the examples to see the framework in action.