✨ LinkedIn Carousel Generator

A complete application that automatically generates professional, highly visual LinkedIn carousels from simple text or articles using Google's Gemini 3 Pro (for advanced reasoning and content structuring) and Gemini 3 Image (for high-quality, style-consistent backgrounds).

Built with Python, Streamlit, and the latest Google GenAI SDK.

🚀 Features

Automated Content Generation: Extracts key points from any text or article and structures them into a compelling multi-slide narrative optimized for LinkedIn.
AI Image Generation: Dynamically creates unique, high-quality background images tailored to the content of each slide using Gemini's image models.
Consistent Styling: Supports multiple visual styles or a custom user-uploaded reference image to maintain brand consistency across all slides.
Auto-Formatting: Automatically overlays text, headlines, and code snippets onto the generated backgrounds using Python Pillow.
Cost Estimation: Provides real-time token and cost estimation for transparency.
Multi-Language Support: Generate carousels in over 10 different languages.
Export Ready: Download all generated slides as a ZIP file, ready to be converted to PDF and posted directly to LinkedIn.

🛠️ Tech Stack

Frontend: Streamlit
AI Models: Google Gemini (via google-genai SDK)
Image Processing: Pillow (PIL)
Deployment: Docker & Docker Compose

📦 Getting Started

Prerequisites

A Google Gemini API Key. You can get one from Google AI Studio.
Docker and Docker Compose (Recommended) or Python 3.10+ (for local development).

Option 1: Running with Docker (Recommended)

The easiest way to run the application is using Docker Compose.

Clone the repository (if you haven't already) and navigate to the project directory:
```
cd linkedin-carousel-generator
```
Run the application: You can either export your API key first or pass it inline:
```
export GOOGLE_API_KEY="your_api_key_here"
docker compose up --build -d
```
Alternatively, you can provide the API key later through the web UI.
Access the app: Open your browser and navigate to http://localhost:8501.

Option 2: Local Python Setup

Navigate to the app directory:

cd linkedin-carousel-generator/linkedin_carousel_app

Create a virtual environment:

python -m venv venv
source venv/bin/activate  # On Windows, use `venv\Scripts\activate`

Install the dependencies:
```
pip install -r requirements.txt
```
Run the application:
```
streamlit run app.py
```

🏗️ Project Structure

linkedin-carousel-generator/
├── Dockerfile                  # Container definition for the app
├── docker-compose.yml          # Compose file for easy deployment
├── .dockerignore               # Files to exclude from Docker builds
├── linkedin_carousel_app/      # Main application source code
│   ├── app.py                  # Main Streamlit UI and execution logic
│   ├── generator.py            # Integration with Google Gemini for Text & Images
│   ├── styles.py               # Configuration for visual styles and layouts
│   ├── utils.py                # Helper functions (Image formatting, fonts, etc.)
│   └── requirements.txt        # Python dependencies
└── README.md                   # Project documentation

💡 How to Use

Enter API Key: Open the settings sidebar and paste your Gemini API Key if you haven't set it via environment variables.
Paste Text: In the main area, paste the article, blog post, or ideas you want to convert into a carousel.
Configure: Select an output language and visual style (e.g., Tech Minimalist, Bold Gradient).
Generate: Click "Generate Carousel". The app will first break down your text into slides, and then generate imagery for each one concurrently.
Download: Once finished, preview the slides in the app and click the download button to get a ZIP archive of your completed carousel.

(Tip: To post on LinkedIn as a swipeable carousel, convert the downloaded images into a single PDF document first).

4.4 KiB Raw Blame History