infery
← All models

Iris-Z Text to RGB-D

iris-z-text-to-rgbd

Image generationby Speridlabs

Supports preset or custom sizes up to 1.31 MP with adjustable steps and guidance.

images

Details

Accepts
text

Pricing

Price
17.5 cr / megapixel

Prices in credits (1 credit = $0.01).

Data schema

Input

FieldTypeDescription
seed—The same seed and the same prompt given to the same version of the model will output the same image every time.
promptrequiredstringThe prompt to generate an image from. Prompts are truncated to 512 text-encoder tokens.
image_size—The size of the generated image. Both sides must be multiples of 16 between 256 and 2048, with at most 1310720 pixels. The model was trained at 512 and 1024.
output_formatstringenum: png, jpegThe format of the generated image.
guidance_scale—Classifier-free guidance scale. Defaults to the model's own value (3.5 for Iris-Z, 4.0 for Iris-Z small).
num_inference_stepsintegerThe number of inference steps to perform.
enable_safety_checkerbooleanIf set to true, the safety checker will be enabled. Disabling it requires account authorization; unauthorized requests are always checked, and images flagged as unsafe are returned as black images.

Output

FieldTypeDescription
seedintegerThe seed used for generation.
image—The generated image.
depth_map—Relative log depth as a float32 `.npy` array (height x width). Larger values are farther; sky pixels are 1.0 when the model detects sky.
depth_image—Colorized depth of the generated image (near = red, far = blue).
has_nsfw_conceptsarrayWhether the generated image contains NSFW concepts.