# welcome_text.py

WELCOME_INTRO = """
# 📄 Welcome to the Multimodal Chat-PDF Playground

This space is designed to let you **chat with your PDFs** using both their text and images.

**Key features**  
- 🔍 **OCR & Native Text** – choose whether to run OCR or use the PDF’s embedded text  
- 🖼️ **Image Descriptions** – swap between different vision-language models to caption images  
- 💬 **Chat with Context** – pick from a variety of open-access LLMs (Zephyr, Mistral, Falcon, etc.)  

**Getting Started**  
1. Click **Start** below.  
2. Upload one or more PDFs.  
3. Choose your OCR & vision-language backends.  
4. Extract—wait a few seconds.  
5. Head over to the **Chat** tab and ask anything!

> Feel free to experiment: swap OCR models, try different Llava versions, or test Mistral vs. Falcon for chat.  
> Have fun! 🚀
"""