Summary

Anionex/banana-slides is an AI-native PPT generator democratizing creation. It targets all users, enabling rapid generation of professional, visually appealing slideshows with minimal design effort. The core benefit is empowering users to quickly translate ideas into polished presentations, overcoming the limitations of traditional AI PPT tool limitations.

How It Works

Leveraging the "nano banana pro🍌" model, it generates high-quality, stylistically consistent slides. Supports flexible inputs: ideas, outlines, or page descriptions. Intelligently parses uploaded documents (docx, pdf, md, txt) for content/images and accepts custom materials for style reference. Natural language modification for specific slide elements offers a more intuitive "Vibe PPT" experience than rigid generators.

Quick Start & Requirements

Recommended installation: Docker Compose for simplified frontend/backend deployment. Requires Google Gemini API key in .env. Source deployment needs Python 3.10+ and Node.js 16+. Accessible at http://localhost:3000 (frontend) and http://localhost:5000 (backend API).

Highlighted Details

• Multi-Input Generation: Create presentations from ideas, outlines, or page descriptions.
• Intelligent File Parsing: Extracts content and images from docx, pdf, md, txt files.
• Natural Language Editing: Allows "verbal" modifications to slides and specific areas.
• Custom Material Integration: Upload reference images, screenshots, or existing PPTs for style guidance.
• One-Click Export: Generates PPTX and PDF outputs (16:9 aspect ratio).

Maintenance & Community

Active development: image element segmentation, network search, Agent mode in progress. Future plans include online playback, animations, multi-language support, and user system. "PRs Welcome" indicated; no specific community channels listed.

Licensing & Compatibility

MIT License: permissive for commercial use and integration into closed-source projects.

Limitations & Caveats

Version v0.1.0 suggests early development. Advanced features (image editing, network search, Agent mode) are in progress. Core functionalities (online playback, animations, multi-language) are planned. Requires Google Gemini API dependency.