MemesPy

MemesPy

FastAPI service that generates memes using DSPy for text and OpenAI for images.

How It Works

1. User submits a topic and optional format via the REST API.
2. DSPy (ChainOfThought) generates meme text and an image prompt.
3. Image is generated via gpt-image-1 (saved locally) or DALL-E 3 (URL returned).
4. Meme is stored in SQLite with metadata (generation time, cost, quality scores).

Tech Stack

Component	Technology
API	FastAPI, served by Uvicorn
Text generation	DSPy with OpenAI models
Image generation	gpt-image-1 (primary), DALL-E 3 (fallback)
Database	SQLAlchemy + SQLite
Cache	Redis (optional)
Monitoring	Grafana dashboards, Prometheus

API Endpoints

Method	Path	Description
GET	/api/health	Health check
POST	/api/v1/memes/	Generate a meme
GET	/api/v1/memes/	List memes
GET	/api/v1/memes/{id}	Get a specific meme
GET	/api/v1/analytics/stats/{period}	Generation statistics
GET	/api/v1/analytics/trending	Trending memes
POST	/api/v1/analytics/search	Search by metadata

Requirements

• Python 3.10+
• OpenAI API key

Setup

git clone https://github.com/jmanhype/MemesPy.git
cd MemesPy
python -m venv venv && source venv/bin/activate
pip install -r requirements.txt
cp .env.example .env  # add OPENAI_API_KEY
python scripts/init_db.py
python -m uvicorn src.dspy_meme_gen.api.main:app --port 8081

API docs at `localhost:8081/docs`.

Tests

pytest

Test suite includes unit tests for agents (factuality, image renderer, prompt generator, router, scorer), integration tests, and actor supervision tests.

Status

Working prototype. Meme text generation and image generation function correctly with valid API keys. The analytics and metadata tracking endpoints are built. The repository contains 40+ generated meme PNGs committed to `static/images/memes/` which inflates the repo size. The arti