sligter/LandPPT

一个基于LLM的演示文稿生成平台,能够自动将文档内容转换为专业的PPT演示文稿。平台支持多种AI模型,提供丰富的模板和样式选择,让用户能够创建高质量的演示文稿。

What it solves

LandPPT is an AI-powered platform that automates the entire process of creating professional presentations. It eliminates the manual effort of drafting outlines, designing layouts, finding relevant images, and writing speaker notes, transforming a simple topic or a document into a complete, editable presentation.

How it works

The platform uses a structured AI workflow:

  1. Input: Users provide a topic description or upload documents (PDF, Word, Markdown, Excel, PPT).
  2. Planning: The AI generates an editable outline, which can be enhanced with real-time web research using Tavily or SearXNG.
  3. Generation: The system produces HTML-based slides with AI-adaptive layouts and templates.
  4. Refinement: Users can edit content and styles via an AI chat sidebar and swap images using integrated image services.
  5. Delivery: The final product can be exported to various formats (PDF, HTML, PPTX, DOCX, Markdown) or shared as a public link with accompanying audio narration.

Who it’s for

It is designed for anyone needing to create high-quality presentations quickly, from students and educators to business professionals, as well as developers who want to integrate PPT generation via a REST API.

Highlights

  • Multi-AI Provider Support: Compatible with OpenAI, Anthropic, Google Gemini, Azure OpenAI, and local models via Ollama.
  • Deep Research Integration: Real-time information gathering and summarization using Tavily and SearXNG.
  • Intelligent Imaging: Combines local libraries, web sources (Pixabay, Unsplash), and AI generation (DALL-E, SiliconFlow).
  • Comprehensive Export Options: Supports standard editable PPTX (via Apryse), image-based PPTX, PDF, and 1080p presentation videos with Edge-TTS audio.
  • Flexible Deployment: Offers simple local startup via uv or pip, and a production-ready Docker Compose stack including PostgreSQL, Valkey, and MinIO.

Related

  • Project
  • Project
  • Project
  • Project
  • Project