AzureMicrosoft image model

MAI-Image-2.7. Microsoft AI images, precision or Flash.

Microsoft AI's newest image family: MAI-Image-2.7 for design-ready stills and MAI-Image-2.7 Flash for speed and volume. Both generate from a prompt or edit from up to five references.

  • Text and image input
  • MAI-Image-2.7 and Flash
  • 7 aspect ratios plus auto
  • Up to 5 reference images
  • Transparent background

MAI-Image-2.7 is Microsoft AI's image generation and editing family. MAI-Image-2.7 is the precision tier; MAI-Image-2.7 Flash is the lower-cost tier for high-volume work. Both take a prompt and up to five reference images, in seven aspect ratios plus auto, and can return a transparent background.

In our tests both returned 1024 pixel squares. MAI-Image-2.7 billed about 3.5 cents an image and Flash about 1 cent, and an edit with one reference kept the product in its new scene on both, in 11 to 12 seconds. On Treza both sit in the Image Generation node's picker and in chat.

Two tiers, one prompt

Draft on Flash at about a cent an image, then re-run the keeper on MAI-Image-2.7. The prompt carries across unchanged.

Up to five references

Put a product into a new set or combine several images into one. Our one-reference edits kept the product and changed the scene.

Transparent backgrounds

Ask for a transparent background to get a cut-out ready to drop onto a thumbnail, a slide or a video layer.

Seven aspect ratios

1:1, 4:3, 3:4, 16:9, 9:16, 3:2 and 2:3, plus auto, for feeds, stories and thumbnails.

What to make with MAI-Image-2.7

Real prompts, and what comes back. Every one runs as a node on the canvas, and most run in chat.

“A ceramic cup of coffee on a sunlit wooden cafe table, steam rising, shallow depth of field”

A sunlit cafe still with natural light.

“The same folded shirt (attached) on a white marble counter in a bright studio”

Your product in a new set.

“A sneaker on a transparent background, three-quarter view, soft studio light”

A cut-out ready to layer.

“Six variations of a minimalist candle on stone, on Flash”

Quick options to pick a direction.

“A 16:9 hero image of a modern kitchen at golden hour for a landing page”

A wide, design-ready header image.

“A 9:16 portrait of a florist arranging peonies in her shop”

A vertical lifestyle still for stories.

MAI-Image-2.7 is one node on a canvas

Chain it with language, image, and video models, then publish the whole pipeline as an API. See how a full AI video pipeline fits together.

  1. Step 01

    Draft the brief

    A language model node expands a one-line idea into a detailed, on-brand prompt before it ever reaches the model.

  2. Step 02

    Generate with MAI-Image-2.7

    Run variants in parallel, compare against other models on the same prompt, and keep every result in run history.

  3. Step 03

    Publish it as an API

    The pipeline becomes a versioned endpoint. Your product calls it with the typed /invoke API or any OpenAI SDK.

MAI-Image-2.7 as an API, priced per image

The MAI-Image-2.7 API

Build a pipeline with MAI-Image-2.7 on a node and publish it. It becomes a versioned HTTP endpoint: the typed /invoke route enqueues a run and returns a run id to poll, and an OpenAI-compatible route works with SDKs you already use. Agents connect over MCP the same way. Your prompt engineering ships inside the pipeline, so every caller gets your quality, and swapping the model later does not change the API your product calls.

MAI-Image-2.7 pricing

Prepaid credits, no subscription. Each image is charged at MAI-Image-2.7's own rate from your balance, shown before you run, and a failed generation costs nothing. Credit packs start at $5 and never expire, which suits spiky production schedules. Run history records the cost of every generation, per node, so spend is auditable rather than a surprise.

MAI-Image-2.7, answered

What is the difference between MAI-Image-2.7 and MAI-Image-2.7 Flash?

They take the same inputs: a prompt, up to five reference images, seven aspect ratios plus auto, and an optional transparent background. MAI-Image-2.7 is the precision tier and Flash the cheaper one: about 3.5 cents against about 1 cent an image in our tests.

What does MAI-Image-2.7 cost per image?

OpenRouter lists it per image token, and each image in our tests was 1024 output tokens: about 3.5 cents on MAI-Image-2.7 and about 1 cent on Flash. A reference image added about 1.3 cents on MAI-Image-2.7 and 0.4 cents on Flash. Treza quotes those amounts, plus its credit markup, before a run.

How does it compare with MAI-Image-2.6?

The inputs match, with transparent backgrounds added. Per image, MAI-Image-2.7 lists slightly below MAI-Image-2.6 and its Flash tier at about half of 2.6 Flash. Run one prompt through both on two nodes and keep the one you prefer.

Can I use it in chat?

Yes. Ask for MAI-Image-2.7 or MAI-Image-2.7 Flash by name in chat, attach reference images if you have them, and the image comes back in the conversation.

How much does MAI-Image-2.7 cost on Treza?

Treza uses prepaid credits with no subscription. Each image is charged at MAI-Image-2.7's rate from your balance, and only successful generations are charged. Credit packs start at $5 and never expire.

Can I call MAI-Image-2.7 through an API?

Yes. Build a pipeline that uses MAI-Image-2.7 and publish it as a versioned HTTP endpoint. Call the typed /invoke API from your product, or point any OpenAI SDK at the OpenAI-compatible endpoint. If a better model ships later, swap it on the node without changing the API your product calls.