dezgo-img-gen

Dezgo Image Generation

Generate images via the local freegen container at http://freegen:3000.

POST /generate

Required fields: prompt, category (or model for flux/backwards compat)

Examples

Flux (fast, no model selection needed):

{"prompt": "a cute cat", "model": "flux", "width": 1024, "height": 1024}

SDXL (with model selection):

{"prompt": "portrait of a woman", "category": "sdxl", "model": "juggernautxl_1024px", "width": 1024, "height": 1024}

IMG2IMG:

{"prompt": "red dress", "category": "i2i", "model": "realistic_vision_5_1", "init_image": "data:image/png;base64,..."}

Image tools

• RemoveBG: {"category": "removebg", "init_image": "data:image/png;base64,..."}
• Upscale: {"category": "upscale", "init_image": "data:image/png;base64,...", "scale": 2}
• Inpainting: {"category": "inpaint_sdxl", "prompt": "filling item", "init_image": "data:image/png;base64,...", "mask_image": "data:image/png;base64,..."}
• Text Inpainting: {"category": "textinpaint", "mask_prompt": "shirt", "prompt": "blue shirt", "init_image": "data:image/png;base64,..."}

General options

• width, height – image dimensions
• steps – quality vs speed (flux: 4, sdxl: 20-30, sd1: 20-30)
• guidance – prompt adherence (default 7, sdxl2 lightning: 2)
• seed – reproducible results
• sampler – auto, dpmpp_2m_karras, euler, etc.

Response

Returns JSON: {success, file, url, size, elapsed, cost, balance, model, sub_model}

GET /info

Lists available models for each category.

GET /health

Container health check.