What is PaperBanana?

PaperBanana is an agentic framework for automated generation of publication-ready academic illustrations. It uses a multi-agent workflow with specialized agents—Retriever, Planner, Stylist, Visualizer, and Critic—to transform raw scientific content into professional diagrams and plots. It supports methodology diagrams, statistical plots, aesthetic enhancement, educational infographics, and aesthetic refinement.

How to use PaperBanana?

Input your research content—methodology, data, or concepts—via text descriptions or upload reference images for style guidance. Select the illustration type, aspect ratio, and resolution, then click 'Generate Images'. Review the output and refine prompts or use the Aesthetic Refinement feature to polish specific visual elements. Download high-resolution images or generated Python code for statistical plots.

Core features of PaperBanana

Multi-Agent Workflow: Five specialized agents (Retriever, Planner, Stylist, Visualizer, Critic) collaborate to transform content into polished illustrations.
Reference-Driven Style: Retrieves relevant academic references to guide visual style and match publication standards.
Iterative Refinement: Critic agent automatically reviews and refines images until quality standards are met.
Code-Based Statistical Plots: Generates executable Python Matplotlib code for accurate data visualization.
Diverse Illustration Types: Supports methodology diagrams, statistical plots, aesthetic enhancement, educational infographics, and aesthetic refinement.
Publication-Ready Output: High-resolution images optimized for research papers, presentations, and academic posters.