Microsoft Opens MAI Image 2.7 Preview for Image Editing
Microsoft says the model improved on MAI Image 2.6 across all 10 tested use cases and seven editing actions. The company has not set a date for text-to-image generation.

Microsoft introduced MAI Image 2.7 on Oct. 27 as a preview model for editing existing images. The company has not announced when its planned text-to-image generation feature will be released.
The model can restyle or relight scenes, change text and symbols within images, and restore or enhance existing images. Microsoft says it improved on MAI Image 2.6 across all 10 tested use cases and seven editing actions. The company also reports a 24-Elo lead over the next model on scene and style tasks.
An image-editing leaderboard places MAI Image 2.7 at No. 2, with 1,193 Elo across 4,891 samples. Its 95% Elo interval is 1,178 to 1,208. The ranking uses pairwise human-preference votes, and the model’s interval overlaps the first-ranked model’s interval, so the placement alone does not establish a clear performance gap.
The leaderboard does not list an API as available. Microsoft has not disclosed pricing or general availability details. The company described text-to-image generation as coming soon but gave no release date.