Imagen

Generate, upscale, and edit images using Imagen models

Imagen models enable you to generate, upscale, and edit images using the Google Gen AI Python SDK.

Generate images

Generate images from text prompts using the generate_images method:

from google.genai import types

response = client.models.generate_images(
    model='imagen-4.0-generate-001',
    prompt='An umbrella in the foreground, and a rainy night sky in the background',
    config=types.GenerateImagesConfig(
        number_of_images=1,
        include_rai_reason=True,
        output_mime_type='image/jpeg',
    ),
)
response.generated_images[0].image.show()

Configuration options

The GenerateImagesConfig supports the following parameters:

• number_of_images - Number of images to generate (default: 1)
• include_rai_reason - Include Responsible AI filtering reasons
• output_mime_type - Output format: image/jpeg or image/png

Upscale images

Image upscaling is only supported in Vertex AI.

Upscale existing images to higher resolutions:

from google.genai import types

response = client.models.upscale_image(
    model='imagen-4.0-upscale-preview',
    image=response.generated_images[0].image,
    upscale_factor='x2',
    config=types.UpscaleImageConfig(
        include_rai_reason=True,
        output_mime_type='image/jpeg',
    ),
)
response.generated_images[0].image.show()

The upscale_factor parameter accepts 'x2' or 'x4' for 2x or 4x upscaling.

Edit images

Image editing is only supported in Vertex AI and uses a separate model (imagen-3.0-capability-001).

Edit images using inpainting and masking techniques:

from google.genai import types
from google.genai.types import RawReferenceImage, MaskReferenceImage

# Define the reference image
raw_ref_image = RawReferenceImage(
    reference_id=1,
    reference_image=response.generated_images[0].image,
)

# Define the mask (model computes a mask of the background)
mask_ref_image = MaskReferenceImage(
    reference_id=2,
    config=types.MaskReferenceConfig(
        mask_mode='MASK_MODE_BACKGROUND',
        mask_dilation=0,
    ),
)

response = client.models.edit_image(
    model='imagen-3.0-capability-001',
    prompt='Sunlight and clear sky',
    reference_images=[raw_ref_image, mask_ref_image],
    config=types.EditImageConfig(
        edit_mode='EDIT_MODE_INPAINT_INSERTION',
        number_of_images=1,
        include_rai_reason=True,
        output_mime_type='image/jpeg',
    ),
)
response.generated_images[0].image.show()

Edit modes

The edit_mode parameter supports:

• EDIT_MODE_INPAINT_INSERTION - Insert new content into masked regions
• EDIT_MODE_INPAINT_REMOVAL - Remove content from masked regions
• EDIT_MODE_OUTPAINT - Extend the image beyond its borders

Mask modes

The mask_mode parameter supports:

• MASK_MODE_BACKGROUND - Mask the background
• MASK_MODE_FOREGROUND - Mask the foreground
• MASK_MODE_SEMANTIC - Mask based on semantic segmentation