qwen-image-3 model. Use a prompt alone for text-to-image, or add one to three reference URLs for image editing. GoEnhance provides independent API access; it is not an official Alibaba API or an Alibaba affiliate.Qwen-Image 3.0 API
Generate or edit one image through GoEnhance with the
Create an API Key
Build Generation and Editing Into One Workflow
One Endpoint, Two Modes
POST to /api/v1/images/generations with model: qwen-image-3. Supply only a prompt for text-to-image, or include one to three accessible reference-image URLs for editing. Every request returns one image.
Asynchronous Results by Design
A successful submission returns an img_uuid. Persist it, poll GET /api/v1/jobs/detail, or provide a public HTTPS callback URL to receive processing, success, and failed status changes.
Validate the documented fields before submitting a generation or edit task.
Qwen-Image 3.0 API Parameters
Model ID
Set <code>model</code> to the exact value <code>qwen-image-3</code>.
Prompt
Provide 1–2000 characters describing the desired image or edit.
Aspect Ratio
Use <code>auto</code> or a documented ratio from 1:1 through 1:8 and 8:1.
Output Tier
Choose <code>1K</code> or <code>2K</code>; the documented default is 2K.
Reference Images
Optionally provide one to three image URLs, each under 10 MB, for editing.
Prompt Controls
Optionally set a negative prompt, seed, and direct or agent prompt extension mode.
Pricing is based on the output plus any reference-image surcharge.
Qwen-Image 3.0 API Pricing
Output Price
Each 1K or 2K output costs 1.75 tokens, equal to $0.035 per generated image.
Reference Surcharge
Each reference image adds 0.19 tokens, equal to $0.0039 per reference image.
Task Accounting
A request with two references costs 2.13 tokens: 1.75 + (2 × 0.19).
Integrate the Qwen-Image 3.0 API
1
Create the Request
Send your API credential in the Authorization header, then POST a JSON body with
model and prompt. Add ratio, image size, or reference URLs only when needed. 2
Persist the Task ID
Check for
code: 0 and store the returned img_uuid. Treat insufficient-token and authentication responses as failures rather than image results. 3
Resolve Completion
Poll the job-detail endpoint or use a public HTTPS callback. Accept the asset only after the task reaches success; handle failed status explicitly.
Add image generation and controlled edits to production workflows without building a model-serving stack.
Developer Use Cases
Product Visuals
Create campaign-ready product concepts from structured prompts.
Creative Variations
Generate fresh visual directions for ads, posts, and landing pages.
Reference-Guided Edits
Transform one to three supplied references while preserving prompt intent.
Localized Campaigns
Produce market-specific visual concepts from localized creative briefs.
Asset Pipelines
Queue asynchronous generation jobs inside automated content workflows.
Design Prototyping
Explore composition, material, and color ideas before final production.
Answers to common pricing, input, and integration questions.