Qwen3 Pro Image — Qwen3 Image's premium tier
Qwen3 Pro Image is the higher-tier model of the Qwen3 pair. It publishes the same options as Qwen3 Image: text-to-image and image-to-image, 1k (default) or 2k, eight aspect ratios, and up to 3 reference images.
Highlights
1K / 2K · Text to image · Image to image
Best for
Final renders and polished work in the Qwen3 family after drafting with Qwen3 Image.
Input
Text, up to 3 reference images
Output
Image
Checking your session…
Pricing
Prepaid in US dollars. We hold the quoted amount while it runs and charge it only on delivery — failed runs are free.
Cost of your first test
$0.0328
One image at the lowest-cost settings.
What your top-up buys
Top up$10
≈ 304 test runs
$50
≈ 1521 test runs
$100
≈ 3043 test runs
| Model | Mode | Resolution | Our price | Official | You save |
|---|---|---|---|---|---|
Qwen3 Pro Image Alibaba | Text to image | 1K | $0.0328/img | $0.04/img | -17% |
Qwen3 Pro Image Alibaba | Text to image | 2K | $0.0657/img | $0.075/img | -12% |
Qwen3 Pro Image Alibaba | Image to image | 1K | $0.0531/img | ||
Qwen3 Pro Image Alibaba | Image to image | 2K | $0.096/img |
What is the Qwen3 Pro Image API?
The Qwen3 Pro Image API exposes Alibaba's higher-tier Qwen3 image model. Its modes, resolutions, aspect ratios and reference limit match Qwen3 Image, so moving between them is a model id change. Images are billed per image.
2 ways to generate, one endpoint
Set mode, send the inputs it needs — that's it.
Text to image
mode: "t2i" · prompt
Describe the image you want in the prompt.
Image to image
mode: "i2i" · reference_images (≤ 3)
Add up to 3 reference images, then describe the edit or the new image.
What you can build with this model
Workflows this model is a good fit for.
Final-quality visuals
Re-run drafts from Qwen3 Image on the higher tier before publishing.
Try it in the playgroundEdits guided by reference
Refine existing artwork by passing up to 3 reference images with an updated prompt.
Try it in the playgroundMulti-ratio asset sets
Produce the same concept in 1:1, 16:9 and 9:16 for different placements.
Try it in the playgroundWhy run Qwen3 Pro Image on MakerNice
Quote before you pay
Every run gets an exact quote first, locked for 12 minutes. You're charged that amount — never more.
Failed runs are free
We hold the amount while the model runs and charge only on delivery. Failures and queued cancellations cost nothing.
Switch models with one field
Same request, key, balance and console for all 16 models. Not happy with the result? Change model and run it again.
Which should you pick: Qwen3 Pro Image, Qwen3 Image?
Compare modes, length, resolution and references, then run the same prompt on each.
| Qwen3 Pro Image | Qwen3 Image | |
|---|---|---|
| Modes | Text to image · Image to image | Text to image · Image to image |
| Resolution | 1K / 2K | 1K / 2K |
| Reference images | Up to 3 | Up to 3 |
| Starting price | $0.0328/img | $0.0236/img |
Integrate Qwen3 Pro Image in three steps
Step 1
Create an API key
Sign in, top up and create a key in the console. Set a spend limit per key so nothing runs away.
OpenStep 2
Quote and create
POST the input to /v1/generations/quote with the model field set to qwen3-pro-image, then create the generation with the quote id.
OpenStep 3
Poll the result
GET /v1/generations/{id} until the status is succeeded, then download the output.
OpenDocs/API/QWEN3-PRO-IMAGE
API reference
Copy-paste requests for this model. The full parameter reference is in the docs.
curl https://api.makernice.com/v1/generations/quote \
-H "Authorization: Bearer $MAKERNICE_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "qwen3-pro-image",
"mode": "t2i",
"prompt": "A serene Japanese garden in autumn, red maple leaves on still water, soft mist, painterly style.",
"resolution": "1k"
}'Invalid request
400 invalid_request names each failing field. Fields the mode doesn't use are rejected too. Nothing is held.
Key or balance
401 means a missing or invalid key; 402 means the balance or the key's spend limit isn't enough.
Quote problems
A changed price-affecting field returns 422 quote_mismatch; an expired quote 410; a used one 409. Get a new quote.
Failed generation
A failed generation is never charged — the hold is released in full. Keep the generation id and retry when you're ready.
Frequently asked questions
Common questions about this model.
What is Qwen3 Pro Image?
How many reference images does Qwen3 Pro Image support?
How much does Qwen3 Pro Image cost per image?
Qwen3 Pro Image vs Qwen3 Image: which should I use?
Which aspect ratios can I use?
Can I generate images from the console without code?
More Posts
Related models
Compare all modelsSimilar models worth comparing.
ImageQwen3 Image
-21%Everyday images and quick edits: Alibaba's Qwen3 Image at 1K or 2K, eight aspect ratios, up to 3 references.
ImageSeedream 5.0 Pro
-22%ByteDance's Seedream 5.0 Pro: generate or edit images at 1K or 2K, guided by up to 10 reference images.
ImageSeedream 5.0
-4%Blend up to 14 reference images into one: Seedream 5.0 is built for character sheets, collages and style fusion, priced as a single 2K tier.
● Start building
See what it does with your prompt
Test it in the playground, or grab an API key and ship it today.