Skip to content
Model infoModelsSnapshot: 15 Oct 2026

Alibaba Cloud / Wanx image editing

Wanx Local Repaint

Prompt-guided editing inside a supplied mask, with the surrounding image largely preserved. Published access is a finite free trial.

restrictedClosed weightsImage editing
Official documentation (opens in a new tab)Reviewed 14 Oct 2026
31 of 38 specification fields collected
Output resolution1024 × 1024 (default), 720 × 1280, 1280 × 720See specification notes below
Text input limit7575
Input billingSee billing notesAlibaba Model Studio · Beijing · trial only
Output billingSee billing notesAccess and billing notes below

Specifications

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

Context / input limit
Not collected
Maximum input tokens
Not collected
Maximum output
Not collected
Input modalities
Text, ImageSource for Input modalities (opens in a new tab)

Detailed API requires a prompt, even though the summary card lists only Image.

Reasoning
Not collected
Streaming
Not collected
Knowledge cutoff
Not collected
Parameters
Not collected
Architecture
Composer-based image-generation frameworkSource for Architecture (opens in a new tab)

Provider attribution only. Exact deployed implementation and parameter count are not established by the research model.

Published availability regions
China (Beijing)Source for Published availability regions (opens in a new tab)
Synchronous generation
Not supportedSource for Synchronous generation (opens in a new tab)

HTTP is asynchronous. SDK blocking helpers, where provided, wait for an async task; they do not establish a separate synchronous model endpoint.

Request mode
Async HTTP task with X-DashScope-Async: enable; retrieve results via GET /api/v1/tasks/{task_id}.Source for Request mode (opens in a new tab)
Output link expiry
Task IDs and generated-image links remain available for 24 hours. This does not establish input-retention or training-use policy.Source for Output link expiry (opens in a new tab)
Free usage allowance
500 images. API says 180 days after activation; central pricing says 90 days from latest activation, release or approval. Actual entitlement unresolved.Source for Free usage allowance (opens in a new tab)

The competing 180-day term is from this model’s API guide.

Published price
Trial only; paid usage unavailable.Source for Published price (opens in a new tab)
Usage metering
500 generated images in the published trial; calls stop when quota is exhausted.Source for Usage metering (opens in a new tab)
Maximum prompt characters
75Source for Maximum prompt characters (opens in a new tab)

Prompt characters beyond the limit are truncated. This is not a token count.

Image styles
<auto> (default), <3d cartoon>, <anime>, <oil painting>, <watercolor>, <sketch>, <chinese painting>, <flat illustration>Source for Image styles (opens in a new tab)
Required request fields
prompt, base_image_url and mask_image_url.Source for Required request fields (opens in a new tab)
Image input limits
Base and mask strictly greater than 256 × 256 and less than 4096 × 4096 pixels; each at most 10 MB. Mask must match base dimensions.Source for Image input limits (opens in a new tab)
Input image formats
JPG, JPEG, PNG, BMPSource for Input image formats (opens in a new tab)
Image URL requirements
Public HTTP(S) base/mask URLs without Chinese characters; Base64 image data is unsupported.Source for Image URL requirements (opens in a new tab)
Mask behavior
Default empty mask_color binarizes the mask: white marks the area to edit. With RGB color lists, those colors identify edits; other colors are background. This is the opposite of a product-preservation mask.Source for Mask behavior (opens in a new tab)
Output resolutions
1024 × 1024 (default), 720 × 1280, 1280 × 720Source for Output resolutions (opens in a new tab)
Images per request
1–4 outputs; default 1Source for Images per request (opens in a new tab)

Access & pricing

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

Alibaba Model Studio · Beijing · trial only

Provider source (opens in a new tab)
API model ID
wanx-x-painting
Access
POST https://{WorkspaceId}.cn-beijing.maas.aliyuncs.com/api/v1/services/aigc/image2image/image-synthesis. Async HTTP task with X-DashScope-Async: enable; retrieve results via GET /api/v1/tasks/{task_id}.
Regions
China (Beijing)
Published request quota
120 requests per minuteSource for Published request quota (opens in a new tab)
Published token quota
1,000,000 tokens per minuteSource for Published token quota (opens in a new tab)

Provider card quota only; not a context limit or image billing unit.

Task submission rate
2 task submissions per secondSource for Task submission rate (opens in a new tab)
Quota scope
Per-model limits shared by Alibaba Cloud account and RAM usersSource for Quota scope (opens in a new tab)
Free usage allowance
500 images. API says 180 days after activation; central pricing says 90 days from latest activation, release or approval. Actual entitlement unresolved.Source for Free usage allowance (opens in a new tab)

The competing 180-day term is from this model’s API guide.

Request mode
Async HTTP task with X-DashScope-Async: enable; retrieve results via GET /api/v1/tasks/{task_id}.Source for Request mode (opens in a new tab)
Synchronous generation
Not supportedSource for Synchronous generation (opens in a new tab)

HTTP is asynchronous. SDK blocking helpers, where provided, wait for an async task; they do not establish a separate synchronous model endpoint.

Endpoint pricing has not been collected. Check the provider for applicable rates.

Data & enterprise terms

Generated links expire after 24 hours; input retention, training use and account-specific access remain unverified.

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.

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

Classification inference: separately named native Alibaba hosted release; the reviewed API and card expose service access without model weights. Its referenced Composer research repository contains documentation and demonstration assets, with weights still listed as a release TODO. Hosted checkpoint lineage is unverified; public Wan releases and style presets are not assumed equivalent.

Classification evidence (opens in a new tab)

Lifecycle dates

Model Studio release28 May 2024Source (opens in a new tab)

Sources reviewed