FluxCode is a sophisticated AI-powered code assistant that leverages Google's Gemini AI to help developers with coding tasks, debugging, explanations, and more. Built with a modern, professional dark-theme interface using Streamlit, it offers an intuitive chat-based experience for all your programming needs.

Feature Description Modern UI Professional dark theme with smooth animations and glassmorphism effects Smart AI Powered by Google's latest Gemini 2.0 Flash and 1.5 Pro models Persistent Sessions Save, load, and export conversations with full metadata Multiple Modes Combine code generation, debugging, and explanation modes Session Analytics Track messages and session duration in real time Multi-turn Chat Full conversation history passed to the model for context-aware responses

Code Generation Mode — Generates clean, well-commented code for any task

Explanation Mode — Detailed conceptual breakdowns for learning

Debug Mode — Targeted debugging assistance with issue identification

Response Styles — Choose Concise, Balanced, or Detailed verbosity

Model Best For gemini-2.0-flash Fast responses, everyday coding tasks gemini-1.5-pro Complex reasoning, large context windows gemini-1.5-flash Balanced speed and capability

Save conversations with custom titles

Load any previously saved session

Export conversations as JSON

Auto-save support

Python 3.8+

Google Gemini API key — get one free at Google AI Studio

Then open http://localhost:8501 in your browser.

Create a .env file in the project root:

Go to Google AI Studio

Sign in with your Google account

Click Create API key

Paste the key into your .env file, or enter it directly in the sidebar

Launch the app with streamlit run app.py

Enter your Gemini API key in the sidebar

Select your AI mode(s) and response style

Type your question in the chat input and press Enter

Action How To Save Click 💾 Save in the sidebar Load Click any saved conversation title Delete Click 🗑️ next to the conversation Export Click 📁 Download JSON New chat Click 🆕 New (auto-saves if enabled)

You can enable multiple modes simultaneously for richer responses — e.g., enable both Code Generation and Explanation to get code with a walkthrough, or combine Debug and Explanation for annotated bug analysis.

Function Description initialize_session_state() Sets up all session variables with defaults generate_response(prompt, api_key) Calls Gemini API with full conversation history format_response_with_mode(prompt) Prepends system instructions based on active modes save_conversation() Persists current chat with timestamp and metadata load_conversation(conv_id) Restores a saved conversation by ID export_conversation() Serializes current chat to a JSON string create_sidebar() Renders the full sidebar UI and returns the API key display_message(message) Renders a chat message with syntax-highlighted code blocks

Variable Type Description messages List[Dict] Full chat history with role, content, id, timestamp current_model str Active Gemini model identifier conversation_title str Title of the current session saved_conversations Dict All persisted conversations keyed by ID user_preferences Dict Theme, response style, and auto-save settings code_gen_mode bool Code Generation mode toggle explain_mode bool Explanation mode toggle debug_mode bool Debug mode toggle response_style str "Concise" / "Balanced" / "Detailed"

Contributions are welcome! To get started:

Follow PEP 8 style

Write clear, descriptive commit messages

Update this README if you add or change features

Test your changes before submitting

New AI modes or prompt templates

UI/UX improvements

Test coverage

Internationalization (i18n)

Plugin/extension architecture

Problem Solution Invalid API key Verify the key at Google AI Studio and ensure it's active Package conflicts Use a fresh virtual environment Slow responses Switch to gemini-2.0-flash in the sidebar App crashes on start Ensure Python 3.8+ and all dependencies are installed

For other issues, open a GitHub Issue.

Python: 3.8+

RAM: 512 MB minimum (1 GB+ recommended)

Storage: ~100 MB

Network: Required (Gemini API calls)

Browser: Any modern browser (Chrome, Firefox, Safari, Edge)

Code execution sandbox — run and test code directly in the UI

Team collaboration — share conversations with others

Additional themes and customization options

Voice input via speech-to-text

VS Code extension integration

Plugin system for custom AI modes

This project is licensed under the MIT License.

Google AI for the Gemini models

Streamlit for the web framework

All contributors who help improve FluxCode

Made with ❤️ for developers

⭐ Star this repo · 🐛 Report a bug · 💡 Request a feature