Paper2Poster
[NeurIPS 2025] Open-source Multi-agent Poster Generation from Papers
Paper2Poster is an agent skill that transforms scientific papers into professional academic posters through a multi-agent system called PosterAgent. It processes PDF papers through a parser, planner, and painter-commentor loop to generate editable PowerPoint posters with visual-in-the-loop refinement. The skill integrates with AI agents like Codex and Claude, enabling researchers to automate poster creation with simple commands. Developed as part of research accepted to NeurIPS 2025, it supports flexible LLM/VLM combinations including GPT-4o, Qwen, and local models via vLLM.
Key Features
Use Cases
- 01Automating academic poster creation for conferences like NeurIPS or CVPR from research papers
- 02Generating poster drafts within AI coding assistants like Codex or Claude using natural language commands
- 03Creating presentation materials for research groups using local open-source models to avoid API costs
- 04Producing multiple poster variations with different visual styles through YAML configuration
- 05Evaluating poster quality against human-created posters using the Paper2Poster benchmark dataset
- 06Rapid prototyping of poster layouts for thesis defenses or lab meetings
Related Skills
View moresuperpowers
An agentic skills framework & software development methodology that works.
skills
Skills for Real Engineers. Straight from my .agents directory.
skills
Public repository for Agent Skills
ponytail
Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.
Paper2Poster — FAQ
What is Paper2Poster and how does it work with AI agents?+
Paper2Poster is an agent skill that enables AI assistants like Codex or Claude to generate academic posters from PDF papers. The lightweight skill in the skills/ directory can be invoked with commands like 'Use $paper2poster-poster to turn this paper into a poster package' without requiring additional dependencies.
How do I install Paper2Poster for standalone use?+
Install Python dependencies with pip install -r requirements.txt, then install LibreOffice (for poster rendering) and poppler (via conda install -c conda-forge poppler). Create a .env file with your OPENAI_API_KEY for GPT-4o access, or deploy local models using vLLM for API-free operation.
Which AI clients and models does Paper2Poster support?+
The skill works with Codex and Claude agents. The standalone system supports GPT-4o, GPT-4o-mini, Qwen-2.5-7B-Instruct, and other models deployable via vLLM, with flexible LLM/VLM combinations for text and vision tasks.
Do I need API keys or can I run it locally?+
API keys are optional. For cloud-based generation, provide an OPENAI_API_KEY in your .env file. For fully local operation, deploy open-source models like Qwen using vLLM and specify vllm_qwen as the model name.
Is Paper2Poster free and open-source?+
Yes, Paper2Poster is open-source and available on GitHub. While the code is free, using commercial APIs like GPT-4o incurs costs; local deployment with open-source models is completely free.
Can I customize the poster style and add logos?+
Yes, you can customize appearance via YAML configuration files (global config/poster.yaml or per-poster overrides). Logos for conferences and institutions can be added automatically through web search or by specifying custom logo paths.
How do I install Paper2Poster?+
Open the source repository on GitHub and follow its README. Paper2Poster is a skill — MCP Agents Market links you directly to the official repo.
Is Paper2Poster free?+
Paper2Poster is an open-source project hosted on GitHub. Check the repository for its license and any usage requirements.