All models

P-Image

pruna/p-image
Pruna Text-to-ImageImage Generation

Pruna's P-Image is an ultra-fast text-to-image model with automatic prompt enhancement and 2-stage refinement, combining exceptional speed with high-quality output and flexible aspect ratios.

Quick start

# Inspect the price — a plain request returns the 402 challenge:
curl -i https://api.glianalabs.com/v1/infer \
  -H "content-type: application/json" \
  -d '{
    "model": "pruna/p-image",
    "prompt": "<your prompt>"
  }'

# Pay + run in one step with the mppx CLI (create a wallet: npx mppx account create):
npx mppx https://api.glianalabs.com/v1/infer \
  -J '{"model": "pruna/p-image", "prompt": "<your prompt>"}'

Examples

example

Parameters

Input
prompt string required

Text description of the image to generate. The model automatically enhances prompts for better results.

aspect_ratio string optional default: "16:9"

Aspect ratio for the image. Use "custom" with width/height for exact dimensions.

disable_safety_checker boolean optional default: false

Disable safety checker for generated images.

height integer optional

Custom height in pixels (256-1440, multiple of 16). Only used when aspect_ratio="custom".

hf_api_token string optional

HuggingFace API token for accessing private LoRAs. This credential is forwarded verbatim to Pruna. It is only written to gateway request-body logs when the gateway-level collectLogPayload debug flag is explicitly enabled — it never appears in structured analytics logs.

lora_scale number optional default: 0.5

How strongly the LoRA should be applied (-1 to 3).

lora_weights string optional

Load LoRA weights. Supports HuggingFace URLs in the format huggingface.co/<owner>/<model-name>[/<file.safetensors>].

prompt_upsampling boolean optional default: false

Upsample the prompt with an LLM for enhanced results.

seed integer optional

Random seed for reproducible generation.

width integer optional

Custom width in pixels (256-1440, multiple of 16). Only used when aspect_ratio="custom".

Output
image: model output URL