providerOptions.google.generationConfig.thinkingConfig.thinkingLevel.
The image endpoint accepts text and reference images. Video input, search grounding, and multi-turn editing are not exposed by this endpoint.
Seed, temperature, topP, topK, and logprobs are not supported. Output images use the standard 24-hour URL retention.
Model ID
model field.
Supported modes
image_generationimage_edit
Inputs
TextImage
Outputs
ImageText
Capabilities
- Not specified
Additional capabilities
Text-to-imageImage editingReference imagesUp to 4K output
Technical details
Availability and final customer prices can change. Check the Dashboard model catalog before production use.