DemoGPT

What it is

🤖 Create Agents in a Second - with the Best Tools, RAG, Knowledge Graph and Vector Databases

DemoGPT is a library that enables creating, customizing, and using AI agents with various tools, knowledge resources, and model integrations. It includes an AgentHub for tool creation and agent management, along with examples of tools, agents, and RAG-powered workflows.

How it works

• Provides an AgentHub to install tools and create agents (ToolCallingAgent, ReactAgent, etc.).
• Includes tools like TavilySearchTool, WeatherTool, WikipediaTool, PythonTool, and more for composing agent capabilities.
• Demonstrates usage through code examples for creating tools, assembling agents, and invoking agents with queries.
• Shows RAG integration via BaseRAG for document storage and retrieval to support question answering.

Getting started

• Installation instruction for AgentHub:

pip install demogpt

• Example: creating and using a ToolCallingAgent with TavilySearchTool and WeatherTool:

search_tool = TavilySearchTool()
weather_tool = WeatherTool()
llm = OpenAIChatModel(model_name="gpt-4o-mini")
agent = ToolCallingAgent(tools=[search_tool, weather_tool], llm=llm, verbose=True)

• Example: using an agent:

query = "What's the weather like in New York today?"
response = agent.run(query)
print(response)

• RAG usage example for a document-based query:

rag = BaseRAG(
    llm=OpenAIChatModel(model_name="gpt-4o-mini"), 
    vectorstore="chroma",
    persistent_path="rag_chroma",
    index_name="rag_index",
    reset_vectorstore=True,
    embedding_model_name="sentence-transformers/all-mpnet-base-v2",
    filter={"search_kwargs": {"score_threshold": 0.5}}
)

Getting started (AgentHub specifics)

• Install the main package for DemoGPT AgentHub:

pip install demogpt

• Creating a simple agent example:

from demogpt_agenthub.agents import ToolCallingAgent
from demogpt_agenthub.llms import OpenAIChatModel
from demogpt_agenthub.tools import TavilySearchTool, WeatherTool

search_tool = TavilySearchTool()
weather_tool = WeatherTool()
llm = OpenAIChatModel(model_name="gpt-4o-mini")
agent = ToolCallingAgent(tools=[search_tool, weather_tool], llm=llm, verbose=True)

Recent releases

• v1.2.6 (2023-09-28): DemoGPT v1.2.6 🚀 Release Notes — bug fixes and module enhancements.
• v1.2.5 (2023-09-21): DemoGPT v1.2.5 — release notes mention chat-based application generation.
• v1.2.4 (2023-08-17): DemoGPT v1.2.4 — introduction of a customization property for generated code.
• v1.2.3 (2023-08-13): DemoGPT v1.2.3 — enhancements, integrations, and features.
• v1.2.1 (2023-08-13): (listed as a release in the overview)

Traction

• Stars: 1903

Behind the repo

• Maintained by melih-unsal
• Repository: melih-unsal/DemoGPT
• License: MIT

Caveats

• License: MIT
• Language: Python
• Created: 2023-06-07
• Last push: 2026-04-01
• Open issues: 10