Skip to main content
This node generates or edits images through OpenRouter using Microsoft’s MAI-Image-2.6 models. It supports text-to-image generation as well as image-guided editing with up to five reference images, in seven aspect ratios at either 1K or 1.5K resolution.

Inputs

Common Inputs

Mai Image 2.6 and Mai Image 2.6 Flash Inputs

Shared by both model options (microsoft/mai-image-2.6 and microsoft/mai-image-2.6-flash), which expose the same set of parameters.

Reference Inputs

Notes:
  • A maximum of 5 reference images is supported in total across all connected slots; a batched input counts once per image.
  • The prompt must contain at least 1 character after surrounding whitespace is removed, and cannot exceed 20000 characters.
  • Reference images are sent as PNG data and are limited to a total of 2048 x 2048 pixels.
  • The aspect_ratio setting is applied to the output even when reference images are connected.
  • When aspect_ratio is "auto", the resolution setting is ignored. With no reference images the model picks the ratio for text to image and renders at the 1.5K size; with reference images connected the aspect ratio of the first reference image is kept.
  • The seed value does not affect the API result; it only determines whether the node re-runs.

Outputs

This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub

Source fingerprint (SHA-256): d201c18deccd2523041a24427996f51127ca201dfe10fad60c6f768ab79bf852