Skip to content
Model infoModelsSnapshot: 15 Oct 2026

Vidu (ShengShu) / Vidu Image

Vidu Q2 Fast Image

Vidu image generation and editing variant exposed through Alibaba Model Studio in Beijing. Supports text prompts or up to fourteen reference images.

availableClosed weightsImage generationImage editing
Official documentation (opens in a new tab)Reviewed 14 Oct 2026
32 of 41 specification fields collected
Output resolution1KSee specification notes below
Reference images0–14 imagesSee specification notes below
Input billingSee ratesAlibaba Model Studio · Beijing
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
Not collected
Maximum input tokens
Not collected
Maximum output
Not collected
Reasoning
Not collected
Streaming
Not collected
Knowledge cutoff
Not collected
Release date
Not collected
Parameters
Not collected
Architecture
Not collected
Languages
Chinese, EnglishSource for Languages (opens in a new tab)

Published prompt languages; not an exhaustive claim about all text-rendering languages.

API aliases
vidu/viduq2-fast_reference2imageSource for API aliases (opens in a new tab)
Image input limits
Up to 14 reference images, combined size at most 50 MB; input aspect ratios between 1:4 and 4:1. Exact input pixel bounds are not published.Source for Image input limits (opens in a new tab)
Input media transport
Public HTTP(S) image URLs. Base64 is not documented for this hosted route.Source for Input media transport (opens in a new tab)
Required request fields
model plus input.messages containing exactly one message. At least one non-empty text field is required; the first non-empty text is used as prompt, and all image fields are references.Source for Required request fields (opens in a new tab)
Images per request
1Source for Images per request (opens in a new tab)

n accepts only 1. Field table marks n optional, but a failure example calls it required; explicitly supplying n=1 avoids relying on the inconsistent default.

Supported image sizes
1K: 1024*1024, 1K: 768*1376, 1K: 848*1264, 1K: 896*1200, 1K: 928*1152, 1K: 1152*928, 1K: 1200*896, 1K: 1264*848, 1K: 1376*768, 1K: 1584*672Source for Supported image sizes (opens in a new tab)

Only the enumerated presets are documented; default 1024*1024. Tier names do not imply a fixed long-edge dimension.

Seed control
Integer from 0 through 2,147,483,647; 0 or omission chooses a random seed. Repeated seeds are described as yielding similar outputs, not guaranteed identical images.Source for Seed control (opens in a new tab)
Visible watermark
false by default; true adds a watermark. Location is not specified.Source for Visible watermark (opens in a new tab)
Request mode
Async HTTP only; requires X-DashScope-Async: enable, then poll task status. Provider suggests a five-second polling interval.Source for Request mode (opens in a new tab)
Output link expiry
Output-image download URLs expire after 24 hours. Task expiry and input/training retention are not established by this image contract.Source for Output link expiry (opens in a new tab)
Request quotas
300 requests per minute; these six Vidu image variants share five concurrent running tasks per Alibaba Model Studio API key.Source for Request quotas (opens in a new tab)
Published availability regions
China (Beijing)Source for Published availability regions (opens in a new tab)
Free usage allowance
No free quota published for these hosted image models.Source for Free usage allowance (opens in a new tab)
Usage metering
Only image output is billed, per generated image at the selected resolution; listed original prices exclude promotions.Source for Usage metering (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

Provider source (opens in a new tab)
API model ID
vidu/viduq2-fast_reference2image
Access
Activate Vidu in the Model Studio console and use a Beijing API key. POST https://{WorkspaceId}.cn-beijing.maas.aliyuncs.com/api/v1/services/aigc/image-generation/generation; query GET https://{WorkspaceId}.cn-beijing.maas.aliyuncs.com/api/v1/tasks/{task_id}.
Regions
China (Beijing)
Request mode
Async HTTP only; requires X-DashScope-Async: enable, then poll task status. Provider suggests a five-second polling interval.Source for Request mode (opens in a new tab)
Published request quota
300 requests per minuteSource for Published request quota (opens in a new tab)
Concurrent requests
Five running tasks shared by all six Vidu image variants, per Alibaba Model Studio API keySource for Concurrent requests (opens in a new tab)

Prices retain their published currencies and billing units below.

Published billing units · original currency · observed 14 Oct 2026
CategoryUsageRateUnit
Image generationImage generation · 1K1CN¥0.28125image

Rate conditions

  1. Original Beijing CNY rate per output image, excluding limited-time promotions. Only output is billed; no free quota.
Pricing source (opens in a new tab)

Data & enterprise terms

Vidu is identified as inference provider. Output links last 24 hours; hosted training use and input retention 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

Qualified classification inference: native Vidu service/variant with restricted API access; operator terms reserve proprietary service rights and prohibit reverse engineering. No exact downloadable checkpoint is identified in reviewed primary documentation. This is not proof of private weights, or of distinct checkpoint lineage between similarly named hosted variants.

Classification evidence (opens in a new tab)

Sources reviewed