Qwen Image 3.0 is not publicly callable on EvoLink yet. Alibaba lists the official model ID as qwen-image-3.0-pro. Join the waitlist for verified EvoLink pricing, supported parameters, and route rules. You can prepare your API key and shared image task flow with an available model now.
Get verified EvoLink pricing, supported parameters, route rules, and request guidance when public access opens.
We will only email you about Qwen Image 3.0 availability. No spam.
Want faster updates? Join the EvoLink Discord
Qwen Image 3.0 is Alibaba's latest Tongyi Wanxiang image model. Its public materials focus on text-to-image generation, reference-guided editing, information-dense layouts, small text, and multilingual typography. The official model ID is qwen-image-3.0-pro. EvoLink does not publicly offer this route yet, so the page separates confirmed provider facts from pending EvoLink access details.
These are provider-stated capabilities, not verified EvoLink route guarantees. Test them with your own prompts after public access opens.
Capability illustration · not an EvoLink testAlibaba's public materials demonstrate small text and structured, information-dense layouts across 12 languages. Results can vary by prompt, language, layout, and output size.
Capability illustration · not an EvoLink testAlibaba states that prompts can include detailed layout, typography, style, color, object, and knowledge constraints in one brief.
Capability illustration · not an EvoLink testAlibaba describes text-to-image generation and reference-guided editing. EvoLink-supported inputs, outputs, and model-specific fields will be published at public access.
Official provider facts and EvoLink availability are shown separately so you can distinguish a published model ID from a callable EvoLink route.
Published in Alibaba Cloud Model Studio documentation.
Upstream access currently requires provider approval.
The model is not publicly callable on EvoLink yet.
Official documentation covers generation and reference-guided editing.
The official image-to-image API accepts one to three reference images.
Official output format and image-count range per request.
Official output pixel range for text-to-image and image-to-image tasks.
Pricing will use the existing price area after route and billing verification.
Use these jobs as evaluation targets after public access opens. They do not describe currently callable EvoLink functionality.
Turn a detailed brief into worksheets, exam pages, explainers, slide-like graphics, and structured knowledge visuals with clear hierarchy.
Produce campaign variants with native typography for multilingual ads, ecommerce banners, menus, and launch assets.
Prototype web pages, game screens, livestream interfaces, product packaging, and text-heavy creative directions.
Use reference images to evaluate subject, composition, style, and transformation control in the same image workflow.
Use a fixed prompt suite and compare every result with an available image-model baseline. Provider examples are evidence to test, not a production guarantee.
Test: Use exact copy across English, Chinese, Japanese, and one right-to-left language.
Check: Count missing, changed, duplicated, or unreadable characters.
Test: Specify hierarchy, positions, spacing, and required objects in a repeatable brief.
Check: Score instruction compliance separately from visual appeal.
Test: Render labels, menus, captions, and dense document layouts.
Check: Inspect at delivery size; treat the provider's 10px example as a test target, not a guarantee.
Test: Reuse the same subjects, products, and brand elements across multiple edits.
Check: Measure identity, color, composition, and unwanted changes.
Test: Repeat a fixed prompt set and record completed, failed, and timed-out tasks.
Check: Compare median and tail latency after EvoLink access opens.
Test: Count only outputs that pass your team's quality threshold.
Check: Compare successful-output cost after EvoLink pricing is verified.
Complete the model-agnostic integration now, then verify Qwen Image 3.0-specific behavior when the public EvoLink route opens.
The official model ID is known. EvoLink route mapping, pricing, billing unit, regional availability, rate limits, and supported parameters remain subject to public verification.
Create an EvoLink API key and validate your shared image task flow with an available model now. When Qwen Image 3.0 opens publicly, EvoLink will publish verified route mapping, pricing, billing, and supported parameters.
Not yet. Qwen Image 3.0 is not publicly callable on EvoLink. Join the waitlist on this page to receive verified access details when public availability opens.
Alibaba Cloud Model Studio lists the official model ID as qwen-image-3.0-pro. This does not mean the model is already callable on EvoLink; the EvoLink route mapping will be confirmed at public access.
Yes. Create an EvoLink API key and use an available image model to validate authentication, task submission, polling, storage, errors, and your application flow.
Alibaba Cloud documentation currently lists qwen-image-3.0-pro as an invitation-only preview. Provider access and EvoLink public availability are separate statuses.
Alibaba's public materials focus on text-rich, information-dense images, prompts up to 4.5K tokens, small-text rendering, multilingual typography, and reference-guided editing.
Alibaba's public materials describe reference-guided editing. EvoLink will publish the verified supported inputs, outputs, and model-specific parameters when public access opens.
Alibaba documents text-to-image and image-to-image modes, one to three reference images for editing, PNG output, one to six generated images, and a total output-pixel range from 512×512 to 2048×2048.
The official API reference includes fields such as negative_prompt, prompt_extend, seed, watermark, size, and n. EvoLink's supported subset and request shape will be confirmed at public access.
EvoLink pricing is not publicly confirmed yet. The price and billing unit will appear in the existing pricing area only after the public route and billing are verified.
The verified official sources used for this page describe hosted Model Studio access, not a public open-weight release or license. Treat third-party download or license claims as unverified unless Alibaba publishes them.
Seedream 5.0 Pro, Nano Banana Pro, and GPT Image 2 are available image routes for high-fidelity generation and editing you can evaluate today.
High-fidelity generation and multi-reference creative workflows with flexible output sizes.
Open Seedream 5.0 Pro
Google's high-fidelity general image generation route for flexible creative work.
Open Nano Banana Pro
OpenAI-centered image generation and editing workflows with a documented API.
Open GPT Image 2