AIREITER
Qwen Image 3.0 AI Image Generator and Editor
Generate new images from text or edit up to three references through one production model.
Text to Image
Describe a subject, composition, lighting, and visual style to generate a new image.
Multi-Image Editing
Use one to three reference images for restyling, composition changes, and subject-aware edits.
1K and 2K Output
Choose 1K for faster iteration or 2K when the final image needs more delivered detail.
Up to Six Outputs
Generate one to six images in a single request and receive the exact requested output count.
Verified on the AIReiter OpenAPI
Seven successful production cases
Verified 1:1, 3:2, 2:3, 4:3, 3:4, 16:9, and 9:16 across 1K and 2K output.
Twenty-two delivered images
The standard-model cases covered output_n values from 1 through 6. Every completed task returned the requested image count.
Reference count verified
Text-only, one-image, two-image, and three-image requests all completed successfully.
Billing reconciled
Estimated and final task charges matched for every successful standard-model request. Failed unsupported requests were not charged.
Supported Qwen Image 3.0 Controls
The playground and OpenAPI expose only the controls verified through the connected APIMart channel.
Aspect ratios
1:1, 3:2, 2:3, 4:3, 3:4, 16:9, and 9:16.
Resolution
1K or 2K. Custom pixel dimensions are not exposed on AIReiter for this channel.
Prompt rewrite modes
Prompt extension is off by default. Use direct for text-to-image or image-to-image, and agent for more aggressive text-to-image rewriting.
Negative prompt
Describe unwanted content or artifacts without changing the main prompt.
Qwen Image 3.0 Pricing
Output images
2.40 credits per generated image at either 1K or 2K.
Reference images
Add 0.25 credits for each input image, up to three images per request.
One output with one reference
2.65 credits at either 1K or 2K.
Six outputs with three references
15.15 credits at either 1K or 2K.
Generate with Qwen Image 3.0
Use the same model controls in the browser playground or through the OpenAPI.
Write the prompt
Describe the subject, scene, composition, lighting, typography, and style.
Add references if needed
Upload one to three images for editing or multi-reference composition.
Choose output controls
Select 1K or 2K, one of seven ratios, and one to six outputs.
Generate and compare
Review the delivered images, then change one control at a time for a reliable comparison.
Use Qwen Image 3.0 through the AIReiter API
Submit one normalized task and query the same external task ID until generation reaches a terminal state.
Model ID
Send qwen_image_3 as the model value.
Submit
POST /api/openapi/submit with model, params, and a unique out_task_id.
Query
POST /api/openapi/query with the same out_task_id to read status, consumed credits, and normalized output URLs.
Billing behavior
The submit response shows estimated credits. Completed queries return final credits_used; failed tasks are not charged.
Sources and Verification
Connected supplier
APIMart: https://apimart.ai/zh/model/qwen-image-3-0
Official model reference
Alibaba Cloud Model Studio: https://help.aliyun.com/en/model-studio/qwen-image-generation-and-editing-api-reference
AIReiter API
https://docs.aireiter.com/en/overview
Last verified
August 6, 2026 through https://test-dev.aireiter.com/api/openapi/submit and /api/openapi/query.
Qwen Image 3.0 FAQ
Answers about image generation, editing, sizes, formats, references, pricing, and API usage.
/ 01Can Qwen Image 3.0 generate and edit images?
Yes. Use text only for generation, or add one to three reference images for editing and composition tasks.
/ 02Which aspect ratios are supported?
AIReiter supports 1:1, 3:2, 2:3, 4:3, 3:4, 16:9, and 9:16 through the connected APIMart channel.
/ 03Does AIReiter support custom pixel dimensions?
No. This integration currently exposes validated 1K and 2K presets only.
/ 04Which prompt rewrite modes are available?
Use direct for balanced rewriting in text-to-image or image-to-image requests. Agent is more aggressive and supports text-to-image only.
/ 05How many images can one request return?
Choose one to six outputs. The configured price scales with the requested output count.
/ 06Which model ID should I use?
Use qwen_image_3 with POST /api/openapi/submit, then poll POST /api/openapi/query with the same out_task_id.