AllImageAI

MAI Image 2.5 AI Image Generator

Microsoft’s photoreal image model for natural people, polished materials and design-ready visual concepts.

Developed by Microsoft AIActiveVerified August 25, 2026

Model Version

Mode

0 / 5,000

Preview mode: placeholder output, no credits charged

Output

MAI Image 2.5

Microsoft’s photoreal image model for natural people, polished materials and design-ready visual concepts.

BalancedUp to 1K1:116:99:164:3

Try one of these prompts

Developer
Microsoft AI
Model
MAI Image 2.5
Status
Active
Max Output
1K
Input
Text
Released
June 2, 2026
Last Verified
August 25, 2026

What Is MAI Image 2.5?

MAI Image 2.5 is Microsoft’s photoreal text-to-image model for people, products and polished campaign concepts. The route accepts seven common aspect ratios and keeps the prompt and output format controls deliberately simple.

MAI Image Model Versions

MAI Image 2.5 Pro

Microsoft flagship · 1K

From 9 Credits

Open this version →

MAI Image 2.5

Design-ready · 1K

From 3 Credits

MAI Image 2.5 Pro

Microsoft’s flagship MAI image model for precise typography, rich detail and production-ready compositions.

  • Resolutions: 1K
  • Reference images: not supported
  • Modes: Text to Image, Typography
  • Speed: Slow

MAI Image 2.5

Microsoft’s photoreal image model for natural people, polished materials and design-ready visual concepts.

  • Resolutions: 1K
  • Reference images: not supported
  • Modes: Text to Image, Typography
  • Speed: Balanced

What Is MAI Image 2.5 Best For?

  • Natural lifestyle photography
  • Product and material studies
  • Design-ready concepts
  • Editorial social assets

How to Use MAI Image 2.5

  1. Describe the intended asset and composition.
  2. Put required copy in quotation marks.
  3. Choose the final aspect ratio.
  4. Proofread text and small product details.

MAI Image 2.5 Settings Explained

Aspect Ratio

Choose the delivery composition before generation; the model does not need a separate pixel-size control.

Output Format

PNG is the lossless default; JPEG and WebP are useful for smaller delivery files.

The panel above shows only the parameters MAI Image 2.5 actually supports. Switch versions and unsupported controls disappear rather than sitting there doing nothing.

MAI Image 2.5 Prompt Examples

Campus campaign

Photoreal concept-art poster of a university campus at sunset, cinematic amber light, a fabric banner reading “FUTURE / BUILT”, precise serif typography

Material portrait

Editorial portrait of a ceramic artist in a daylight studio, natural skin texture, chalk dust on linen, glazed vessels catching soft window reflections

MAI Image 2.5 Test Results

Testing in progress

Every model runs the same eight-category prompt set — portrait, product, poster text, anime character, interior, multi-reference, image editing and character consistency — recording generation time, success rate, text accuracy and identity retention. That run has not finished for MAI Image 2.5, so this page shows no scores. We do not publish numbers we have not measured.

MAI Image 2.5 Limitations

  • Prompt tokens add a small variable cost
  • Text-to-image only here
  • One fixed output tier

MAI Image 2.5 Pricing and Credits

Credits scale with resolution and image count. The Generate button always shows what this run costs.

1K

3 credits / image

Failed generations cost 0 credits.

Commercial Use and Image Rights

Commercial use is governed by Microsoft’s model terms and the selected service terms.

  • Model used: MAI Image 2.5
  • Model developer: Microsoft AI
  • The inference route is managed internally and may change without changing the selected model
  • Uploads are sent only to the configured inference service needed to complete the generation
  • You can delete any generation from your history at any time

Frequently Asked Questions

What makes MAI Image 2.5 different?+

It focuses on polished photoreal and design work with a simpler, lower-cost contract than Pro.

Why is the credit price conservative?+

The upstream bill combines image-output tokens with prompt tokens, so credits include room for the full supported prompt length.

Model Updates and Changelog

  • 2026-08-25 —— Added the exact MAI Image 2.5 text-to-image contract with token-aware credits.