PPTAgent

Created By

icip-cas2 months ago

Please run `uv pip install pptagent` first

Content

🌐 Language

PPTAgent: Generating and Evaluating Presentations Beyond Text-to-Slides

📄 Paper | 🤗 OpenSource | 📝 Documentation | DeepWiki | 🙏 Citation

We present PPTAgent, an innovative system that automatically generates presentations from documents. Drawing inspiration from human presentation creation methods, our system employs a two-step process to ensure excellence in overall quality. Additionally, we introduce PPTEval, a comprehensive evaluation framework that assesses presentations across multiple dimensions.

TIP

🚀 Get started quickly with our pre-built Docker image - See Docker instructions

📅 News

[2025/09]: 🛠️ We support MCP server now, see MCP Server for details
[2025/09]: 🚀 Released v2 with major improvements - see release notes for details
[2025/08]: 🎉 Paper accepted to EMNLP 2025!
[2025/05]: ✨ Released v1 with core functionality and 🌟 breakthrough: reached 1,000 stars on GitHub! - see release notes for details
[2025/01]: 🔓 Open-sourced the codebase, with experimental code archived at experiment release

Open Source 🤗

We have released our model and data at HuggingFace.

Demo Video 🎥

https://github.com/user-attachments/assets/c3935a98-4d2b-4c46-9b36-e7c598d14863

Distinctive Features ✨

Dynamic Content Generation: Creates slides with seamlessly integrated text and images
Smart Reference Learning: Leverages existing presentations without requiring manual annotation
Comprehensive Quality Assessment: Evaluates presentations through multiple quality metrics

Case Study 💡

Iphone 16 Pro

Build Effective Agents

PPTAgent 🤖

PPTAgent follows a two-phase approach:

Analysis Phase: Extracts and learns from patterns in reference presentations
Generation Phase: Develops structured outlines and produces visually cohesive slides

Our system's workflow is illustrated below:

PPTAgent Workflow

PPTEval ⚖️

PPTEval evaluates presentations across three dimensions:

Content: Check the accuracy and relevance of the slides.
Design: Assesses the visual appeal and consistency.
Coherence: Ensures the logical flow of ideas.

The workflow of PPTEval is shown below:

PPTEval Workflow

Citation 🙏

If you find this project helpful, please use the following to cite it:

@article{zheng2025pptagent,
  title={PPTAgent: Generating and Evaluating Presentations Beyond Text-to-Slides},
  author={Zheng, Hao and Guan, Xinyan and Kong, Hao and Zheng, Jia and Zhou, Weixiang and Lin, Hongyu and Lu, Yaojie and He, Ben and Han, Xianpei and Sun, Le},
  journal={arXiv preprint arXiv:2501.03936},
  year={2025}
}

Server Config

{
  "mcpServers": {
    "pptagent": {
      "command": "uv",
      "args": [
        "run",
        "--with",
        "pptagent",
        "--with",
        "python-pptx@git+https://github.com/Force1ess/python-pptx@219513d7d81a61961fc541578c1857d08b43aa2a",
        "pptagent-mcp"
      ],
      "env": {
        "PPTAGENT_MODEL": "openai/gpt-4.1",
        "PPTAGENT_API_BASE": "http://localhost:8000/v1",
        "PPTAGENT_API_KEY": "your_key"
      }
    }
  }
}

Recommend Servers

TraeBuild with Free GPT-4.1 & Claude 3.7. Fully MCP-Ready.

MiniMax MCPOfficial MiniMax Model Context Protocol (MCP) server that enables interaction with powerful Text to Speech, image generation and video generation APIs.

Playwright McpPlaywright MCP server

Howtocook Mcp基于Anduin2017 / HowToCook （程序员在家做饭指南）的mcp server，帮你推荐菜谱、规划膳食，解决“今天吃什么“的世纪难题； Based on Anduin2017/HowToCook (Programmer's Guide to Cooking at Home), MCP Server helps you recommend recipes, plan meals, and solve the century old problem of "what to eat today"

EdgeOne Pages MCPAn MCP service designed for deploying HTML content to EdgeOne Pages and obtaining an accessible public URL.

BlenderBlenderMCP connects Blender to Claude AI through the Model Context Protocol (MCP), allowing Claude to directly interact with and control Blender. This integration enables prompt assisted 3D modeling, scene creation, and manipulation.

WindsurfThe new purpose-built IDE to harness magic

Zhipu Web SearchZhipu Web Search MCP Server is a search engine specifically designed for large models. It integrates four search engines, allowing users to flexibly compare and switch between them. Building upon the web crawling and ranking capabilities of traditional search engines, it enhances intent recognition capabilities, returning results more suitable for large model processing (such as webpage titles, URLs, summaries, site names, site icons, etc.). This helps AI applications achieve "dynamic knowledge acquisition" and "precise scenario adaptation" capabilities.

Amap Maps高德地图官方 MCP Server

Visual Studio Code - Open Source ("Code - OSS")Visual Studio Code

Context7Context7 MCP Server -- Up-to-date code documentation for LLMs and AI code editors