paperbanana 0.1.0

Agentic framework for automated academic illustration generation

PaperBanana Automated Academic Illustration for AI Scientists

---

Disclaimer: This is an unofficial, community-driven open-source implementation of the paper "PaperBanana: Automating Academic Illustration for AI Scientists" by Dawei Zhu, Rui Meng, Yale Song, Xiyu Wei, Sujian Li, Tomas Pfister, and Jinsung Yoon (arXiv:2601.23265). This project is not affiliated with or endorsed by the original authors or Google Research. The implementation is based on the publicly available paper and may differ from the original system.

An agentic framework for generating publication-quality academic diagrams and statistical plots from text descriptions. Uses Google Gemini for both VLM and image generation.

• Two-phase multi-agent pipeline with iterative refinement
• Gemini-based VLM planning and image generation
• CLI, Python API, and MCP server for IDE integration
• Claude Code skills for /generate-diagram, /generate-plot, and /evaluate-diagram

---

Quick Start

Prerequisites

• Python 3.10+
• A Google Gemini API key (available at no cost from Google AI Studio)

Step 1: Install

pip install paperbanana

Or install from source for development:

git clone https://github.com/llmsresearch/paperbanana.git
cd paperbanana
pip install -e ".[dev,google]"

Step 2: Get Your API Key

Run the interactive setup wizard:

paperbanana setup

This opens your browser to get a Google Gemini API key from Google AI Studio and saves it to .env.

Or set it up manually:

cp .env.example .env
# Edit .env and add: GOOGLE_API_KEY=your-key-here

Step 3: Generate a Diagram

# Using the included sample input
paperbanana generate \
  --input examples/sample_inputs/transformer_method.txt \
  --caption "Overview of our encoder-decoder architecture with sparse routing"

Or write your own methodology text:

cat > my_method.txt << 'EOF'
Our framework consists of an encoder that processes input sequences
through multi-head self-attention layers, followed by a decoder that
generates output tokens auto-regressively using cross-attention to
the encoder representations. We add a novel routing mechanism that
selects relevant encoder states for each decoder step.
EOF

paperbanana generate \
  --input my_method.txt \
  --caption "Overview of our encoder-decoder framework"

Output is saved to outputs/run_<timestamp>/final_output.png along with all intermediate iterations and metadata.

---

How It Works

PaperBanana implements a two-phase multi-agent pipeline with 5 specialized agents:

Phase 1 -- Linear Planning:

1. Retriever selects the most relevant reference examples from a curated set of 13 methodology diagrams spanning agent/reasoning, vision/perception, generative/learning, and science/applications domains
2. Planner generates a detailed textual description of the target diagram via in-context learning from the retrieved examples
3. Stylist refines the description for visual aesthetics using NeurIPS-style guidelines (color palette, layout, typography)

Phase 2 -- Iterative Refinement (3 rounds):

4. Visualizer renders the description into an image (Gemini 3 Pro for diagrams, Matplotlib code for plots)
5. Critic evaluates the generated image against the source context and provides a revised description addressing any issues
6. Steps 4-5 repeat for up to 3 iterations

Providers

Component	Provider	Model
VLM (planning, critique)	Google Gemini	gemini-2.0-flash
Image Generation	Google Gemini	gemini-3-pro-image-preview

---

CLI Reference

paperbanana generate -- Methodology Diagrams

paperbanana generate \
  --input method.txt \
  --caption "Overview of our framework" \
  --output diagram.png \
  --iterations 3

Flag	Short	Description
--input	-i	Path to methodology text file (required)
--caption	-c	Figure caption / communicative intent (required)
--output	-o	Output image path (default: auto-generated in outputs/)
--iterations	-n	Number of Visualizer-Critic refinement rounds
--vlm-provider		VLM provider name (default: gemini)
--vlm-model		VLM model name (default: gemini-2.0-flash)
--image-provider		Image gen provider (default: google_imagen)
--image-model		Image gen model (default: gemini-3-pro-image-preview)
--config		Path to YAML config file (see configs/config.yaml)

paperbanana plot -- Statistical Plots

paperbanana plot \
  --data results.csv \
  --intent "Bar chart comparing model accuracy across benchmarks"

Flag	Short	Description
--data	-d	Path to data file, CSV or JSON (required)
--intent		Communicative intent for the plot (required)
--output	-o	Output image path
--iterations	-n	Refinement iterations (default: 3)

paperbanana evaluate -- Quality Assessment

Comparative evaluation of a generated diagram against a human reference using VLM-as-a-Judge:

paperbanana evaluate \
  --generated diagram.png \
  --reference human_diagram.png \
  --context method.txt \
  --caption "Overview of our framework"

Flag	Short	Description
--generated	-g	Path to generated image (required)
--reference	-r	Path to human reference image (required)
--context		Path to source context text file (required)
--caption	-c	Figure caption (required)

Scores on 4 dimensions (hierarchical aggregation per the paper):

• Primary: Faithfulness, Readability
• Secondary: Conciseness, Aesthetics

paperbanana setup -- First-Time Configuration

paperbanana setup

Interactive wizard that walks you through obtaining a Google Gemini API key and saving it to .env.

---

Python API

import asyncio
from paperbanana import PaperBananaPipeline, GenerationInput, DiagramType
from paperbanana.core.config import Settings

settings = Settings(
    vlm_provider="gemini",
    image_provider="google_imagen",
    refinement_iterations=3,
)

pipeline = PaperBananaPipeline(settings=settings)

result = asyncio.run(pipeline.generate(
    GenerationInput(
        source_context="Our framework consists of...",
        communicative_intent="Overview of the proposed method.",
        diagram_type=DiagramType.METHODOLOGY,
    )
))

print(f"Output: {result.image_path}")

See examples/generate_diagram.py and examples/generate_plot.py for complete working examples.

---

MCP Server

PaperBanana includes an MCP server for use with Claude Code, Cursor, or any MCP-compatible client. Add the following config to use it via uvx without a local clone:

{
  "mcpServers": {
    "paperbanana": {
      "command": "uvx",
      "args": ["--from", "paperbanana[mcp]", "paperbanana-mcp"],
      "env": { "GOOGLE_API_KEY": "your-google-api-key" }
    }
  }
}

Three MCP tools are exposed: generate_diagram, generate_plot, and evaluate_diagram.

The repo also ships with 3 Claude Code skills:

• /generate-diagram <file> [caption] - generate a methodology diagram from a text file
• /generate-plot <data-file> [intent] - generate a statistical plot from CSV/JSON data
• /evaluate-diagram <generated> <reference> - evaluate a diagram against a human reference

See mcp_server/README.md for full setup details (Claude Code, Cursor, local development).

---

Configuration

Default settings are in configs/config.yaml. Override via CLI flags or a custom YAML:

paperbanana generate \
  --input method.txt \
  --caption "Overview" \
  --config my_config.yaml

Key settings:

vlm:
  provider: gemini
  model: gemini-2.0-flash

image:
  provider: google_imagen
  model: gemini-3-pro-image-preview

pipeline:
  num_retrieval_examples: 10
  refinement_iterations: 3
  output_resolution: "2k"

reference:
  path: data/reference_sets

output:
  dir: outputs
  save_iterations: true
  save_metadata: true

---

Project Structure

paperbanana/
├── paperbanana/
│   ├── core/          # Pipeline orchestration, types, config, utilities
│   ├── agents/        # Retriever, Planner, Stylist, Visualizer, Critic
│   ├── providers/     # VLM and image gen provider implementations
│   │   ├── vlm/       # Gemini VLM provider
│   │   └── image_gen/ # Gemini 3 Pro Image provider
│   ├── reference/     # Reference set management (13 curated examples)
│   ├── guidelines/    # Style guidelines loader
│   └── evaluation/    # VLM-as-Judge evaluation system
├── configs/           # YAML configuration files
├── prompts/           # Prompt templates for all 5 agents + evaluation
│   ├── diagram/       # retriever, planner, stylist, visualizer, critic
│   ├── plot/          # plot-specific prompt variants
│   └── evaluation/    # faithfulness, conciseness, readability, aesthetics
├── data/
│   ├── reference_sets/  # 13 verified methodology diagrams
│   └── guidelines/              # NeurIPS-style aesthetic guidelines
├── examples/          # Working example scripts + sample inputs
├── scripts/           # Data curation and build scripts
├── tests/             # Test suite (34 tests)
├── mcp_server/        # MCP server for IDE integration
└── .claude/skills/    # Claude Code skills (generate-diagram, generate-plot, evaluate-diagram)

Development

# Install with dev dependencies
pip install -e ".[dev,google]"

# Run tests
pytest tests/ -v

# Lint
ruff check paperbanana/ mcp_server/ tests/ scripts/

# Format
ruff format paperbanana/ mcp_server/ tests/ scripts/

Citation

This is an unofficial implementation. If you use this work, please cite the original paper:

@article{zhu2026paperbanana,
  title={PaperBanana: Automating Academic Illustration for AI Scientists},
  author={Zhu, Dawei and Meng, Rui and Song, Yale and Wei, Xiyu
          and Li, Sujian and Pfister, Tomas and Yoon, Jinsung},
  journal={arXiv preprint arXiv:2601.23265},
  year={2026}
}

Original paper: https://arxiv.org/abs/2601.23265

Disclaimer

This project is an independent open-source reimplementation based on the publicly available paper. It is not affiliated with, endorsed by, or connected to the original authors, Google Research, or Peking University in any way. The implementation may differ from the original system described in the paper. Use at your own discretion.

License

MIT