Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.pixeltable.com/llms.txt

Use this file to discover all available pages before exploring further.

​
module  pixeltable.functions.reve

Pixeltable UDFs that wrap Reve image generation API. In order to use them, the API key must be specified either with REVE_API_KEY environment variable, or as api_key in the reve section of the Pixeltable config file.

​
udf  create()

create(
    prompt: String,
    *,
    aspect_ratio: String | None = None,
    version: String | None = None
) -> Image
Creates an image from a text prompt. This UDF wraps the https://api.reve.com/v1/image/create endpoint. For more information, refer to the official API documentation. Parameters:
  • prompt (String): prompt describing the desired image
  • aspect_ratio (String | None): desired image aspect ratio, e.g. ‘3:2’, ‘16:9’, ‘1:1’, etc.
  • version (String | None): specific model version to use. Latest if not specified.
Returns:
  • Image: A generated image
Examples: Add a computed column with generated square images to a table with text prompts: 
t.add_computed_column(img=reve.create(t.prompt, aspect_ratio='1:1'))

​
udf  edit()

edit(
    image: Image,
    edit_instruction: String,
    *,
    version: String | None = None
) -> Image
Edits images based on a text prompt. This UDF wraps the https://api.reve.com/v1/image/edit endpoint. For more information, refer to the official API documentation Parameters:
  • image (Image): image to edit
  • edit_instruction (String): text prompt describing the desired edit
  • version (String | None): specific model version to use. Latest if not specified.
Returns:
  • Image: A generated image
Examples: Add a computed column with catalog-ready images to the table with product pictures: 
t.add_computed_column(
    catalog_img=reve.edit(
        t.product_img,
        'Remove background and distractions from the product picture, improve lighting.',
    )
)

​
udf  remix()

remix(
    prompt: String,
    images: Json,
    *,
    aspect_ratio: String | None = None,
    version: String | None = None
) -> Image
Creates images based on a text prompt and reference images. The prompt may include <img>0</img>, <img>1</img>, etc. tags to refer to the images in the images argument. This UDF wraps the https://api.reve.com/v1/image/remix endpoint. For more information, refer to the official API documentation Parameters:
  • prompt (String): prompt describing the desired image
  • images (Json): list of reference images
  • aspect_ratio (String | None): desired image aspect ratio, e.g. ‘3:2’, ‘16:9’, ‘1:1’, etc.
  • version (String | None): specific model version to use. Latest by default.
Returns:
  • Image: A generated image
Examples: Add a computed column with promotional collages to a table with original images: 
t.add_computed_column(
    promo_img=(
        reve.remix(
            'Generate a product promotional image by combining the image of the product'
            ' from <img>0</img> with the landmark scene from <img>1</img>',
            images=[t.product_img, t.local_landmark_img],
            aspect_ratio='16:9',
        )
    )
)
Last modified on May 1, 2026