Skip to content
Model infoModelsSnapshot: 15 Oct 2026

Microsoft AI / MAI-Image-2.5

MAI-Image-2.5-Flash

MAI-Image-2.5-Flash generates PNG images from text and edits up to five reference images. Available through the Microsoft Foundry image generation and edit APIs.

Image generation
previewClosed weightsImage generationImage editing
Official documentation (opens in a new tab)Reviewed 14 Oct 2026
13 of 22 specification fields collected
Output resolutionAt least 768 pixels per dimension; width × height ≤ 1,048,576 pixels.See specification notes below
Reference imagesUp to 5 images in an edit request, using repeated multipart image fields.See specification notes below
Input billingSee ratesMicrosoft Foundry
Output billingSee ratesPublished units and conditions below

Specifications

Provider-documented facts. “Not collected” means we have not verified that field yet, not that the feature is unsupported.

Context / input limit
32,000 tokensSource for Context / input limit (opens in a new tab)
Maximum input tokens
32,000 tokensSource for Maximum input tokens (opens in a new tab)

Published maximum context length for the prompt.

Maximum output
Not collected
Reasoning
Not collected
Tool calling
Not supportedSource for Tool calling (opens in a new tab)
Structured output
Not collected
Streaming
Not collected
Batch processing
Not collected
Fine-tuning
Not collected
Knowledge cutoff
Not collected
Release date
2026-06-02Source for Release date (opens in a new tab)

Public model introduction from the official release history; distinct from the deployment version.

Parameters
Not collected
Languages
EnglishSource for Languages (opens in a new tab)
Context caching
Not collected
Model revision
2026-06-02Source for Model revision (opens in a new tab)

Provider model version, not a release date.

Input modalities
Text, ImageSource for Input modalities (opens in a new tab)
Output modalities
ImageSource for Output modalities (opens in a new tab)
Input file formats
JPEG, PNGSource for Input file formats (opens in a new tab)
Output formats
PNGSource for Output formats (opens in a new tab)

One base64-encoded PNG in data[].b64_json.

Reference images
Up to 5 images in an edit request, using repeated multipart image fields.Source for Reference images (opens in a new tab)
Output resolutions
At least 768 pixels per dimension; width × height ≤ 1,048,576 pixels.Source for Output resolutions (opens in a new tab)

Either dimension may exceed the equivalent square width as long as the area limit is respected.

Request quotas
Global Standard: free tier 0 has no quota. Tiers 1–6 allow 2, 4, 6, 8, 10, 12 requests/minute respectively.Source for Request quotas (opens in a new tab)

Current documentation snapshot; subscription and deployment tier determine the applicable quota.

Access & pricing

Prices belong to an endpoint and billing tier. A provider listing does not guarantee access for every account or region.

Microsoft Foundry

Provider source (opens in a new tab)
API model ID
Not collected

The model field must contain your Foundry deployment name; the catalog name is not a universal deployment selector.

Access
POST /mai/v1/images/generations with JSON. Catalog model: MAI-Image-2.5-Flash. Authenticate with api-key or Microsoft Entra ID.
Regions
Global Standard deployment; exact resource-region availability not collected.

Prices retain their published currencies and billing units below.

Published billing units · original currency · observed 14 Oct 2026
CategoryUsageRateUnit
Image generationText input1$1.751M tokens
Image generationImage output1$19.501M tokens

Rate conditions

  1. USD list-price snapshot for the us-east pricing selection, per 1M tokens. Global Standard deployment. Region, agreement and taxes may alter charges. Text input, image input and output are separate charges.
Pricing source (opens in a new tab)

Data & enterprise terms

Model-specific retention and training-use terms not collected.

Microsoft Foundry

Provider source (opens in a new tab)
API model ID
Not collected

The model field must contain your Foundry deployment name; the catalog name is not a universal deployment selector.

Access
POST /mai/v1/images/edits with multipart form data. Catalog model: MAI-Image-2.5-Flash. Authenticate with api-key or Microsoft Entra ID.
Regions
Global Standard deployment; exact resource-region availability not collected.

Prices retain their published currencies and billing units below.

Published billing units · original currency · observed 14 Oct 2026
CategoryUsageRateUnit
Image generationText input1$1.751M tokens
Image editingReference image input1$1.751M tokens
Image generationImage output1$19.501M tokens

Rate conditions

  1. USD list-price snapshot for the us-east pricing selection, per 1M tokens. Global Standard deployment. Region, agreement and taxes may alter charges. Text input, image input and output are separate charges.
Pricing source (opens in a new tab)

Data & enterprise terms

Model-specific retention and training-use terms not collected.

Caching

Caching can reuse earlier input processing. Availability, lifetime and billing depend on the endpoint. A cache hit still generates a new response.

01 / INITIAL INPUTProcess & retain

The system processes eligible input and retains reusable state.

02 / LATER REQUESTCheck available state

Reuse depends on matching input, session state and the cache policy.

03 / CACHE HITReuse & generate

Reusing state can reduce repeated work. Any charges follow the billing rules below.

Microsoft Foundry

Not verified

No model-specific cache policy verified for this endpoint.

How it works
Not collected
Minimum reusable input
Not collected
Lifetime & refresh
Not collected
How to enable it
Not collected
Usage fields
Not collected
Billing rules
Cached input rate listed as N/A.Source for Billing rules (opens in a new tab)

This does not prove whether any internal cache exists. No cache support or TTL is inferred.

Limits & data boundaries
Not collected

No verified cache rate for this endpoint. An empty rate is not a zero charge.

Caching reviewed 14 Oct 2026

Microsoft Foundry

Not verified

No model-specific cache policy verified for this endpoint.

How it works
Not collected
Minimum reusable input
Not collected
Lifetime & refresh
Not collected
How to enable it
Not collected
Usage fields
Not collected
Billing rules
Cached input rate listed as N/A.Source for Billing rules (opens in a new tab)

This does not prove whether any internal cache exists. No cache support or TTL is inferred.

Limits & data boundaries
Not collected

No verified cache rate for this endpoint. An empty rate is not a zero charge.

Caching reviewed 14 Oct 2026

What the measurements say

No benchmark observations have been added for this model. We do not turn a vendor description into a performance score.

Latency and throughput depend on the host, request size, reasoning settings and load. There is no measured speed or global usage figure in this record.

The record behind the model

Why it is included

Microsoft identifies this as its own MAI model, supplied through managed inference endpoints. The reviewed model documentation provides no downloadable model weights. This record does not classify public Phi or MAI-DS-R1 weights as closed.

Classification evidence (opens in a new tab)

Sources reviewed