MyAPI Chat

Got an API key? Start chatting in seconds.

Many AI providers offer free tiers or credits to get started:

• Google Gemini — free tier with generous limits
• Groq — free, incredibly fast inference
• OpenRouter — free credits, access to 100+ models
• OpenAI / Anthropic — paid, but industry standards

This app lets you use any of them through a single interface. Paste your key, pick a model, and go. Keys save locally so you don't re-enter them. Chat history persists across sessions.

Built with LangGraph for proper conversation management—easy to extend with RAG, tools, or multi-agent flows later.

Screenshots

Quick start

Requirements: Python 3.12+

git clone https://github.com/SriviharReddy/GenAI_GUI.git
cd GenAI_GUI
uv sync  # or: pip install -r requirements.txt
uv run streamlit run app.py

Paste your API key in the sidebar. Or add to .env:

GEMINI_API_KEY=...
OPENAI_API_KEY=...
ANTHROPIC_API_KEY=...
GROQ_API_KEY=...
OPENROUTER_API_KEY=...

Where to get keys

Provider	Link	Notes
Google Gemini	aistudio.google.com	Free tier available
Groq	console.groq.com	Free, very fast
OpenRouter	openrouter.ai	Free credits, many models
OpenAI	platform.openai.com	Paid
Anthropic	console.anthropic.com	Paid

Project structure

app.py          - Streamlit UI
chatbot.py      - Session & message management
config.py       - Provider configs, sidebar
graph.py        - LangGraph state machine
history.py      - SQLite chat persistence

Extending

Add a new provider in config.py:

PROVIDERS["NewProvider"] = ProviderConfig(
    name="NewProvider",
    env_key="NEW_PROVIDER_API_KEY",
    session_key="new_provider_api_key",
    models=["model-1", "model-2"]
)

Then add the LLM init in graph.py:

case "NewProvider":
    return ChatNewProvider(model=model, api_key=api_key, ...)

Tech

• LangGraph for state management
• Streamlit for UI
• SQLite for chat history persistence
• Supports streaming responses

MIT License