Morph Me

Vision → Generation Pipeline:

Each iteration: Vision AI describes the image, then SDXL generates a new version based on that description + your style. This creates a "telephone game" effect!

The Process:

1. Take your photo
2. Vision AI describes it in detail
3. SDXL generates new image from description + style
4. Repeat 5 times, each using the previous output

Models:

LLaVA 1.5 7B Vision SDXL Lightning Generation

API:

POST /api/gpu/vision/describe POST /api/gpu (SDXL)