Alibaba Qwen Image 2.0 is a unified image model for realistic generation, structured text rendering, native 2K output, and flexible editing across creative and visual design workflows.
The prompt to edit the image with
0/5000Click to upload or drag and drop
Supported formats: JPEG, PNG, WEBP Maximum file size: 10MB; Maximum files: 3
The reference image to guide the edit
Image ratio
Generate image format
The same seed and the same prompt given to the same version of the model will output the same image every time
A configurable parameter. Defaults to true in the Playground.
Explore different use cases and parameter configurations
Describe the content of the generated image
0/5000Image ratio
Generate image format
The same seed and the same prompt given to the same version of the model will output the same image every time
A configurable parameter. Defaults to true in the Playground.
Explore different use cases and parameter configurations
Complete guide to using Qwen Image 2.0
Access Alibaba’s Qwen-Image-2.0 API through Kie.ai for high-quality image generation, advanced text rendering, native 2K outputs, and natural-language image editing.
Qwen Image 2.0 API supports longer and more detailed prompt instructions for text-heavy visuals such as infographics, PPT-style slides, posters, and comics. For workflows that rely on readable in-image text, clearer layout structure, and more controlled visual composition, Qwen Image 2.0 API is a stronger fit for design-focused image generation.
Qwen Image API is built for more detailed and realistic image generation across people, nature, and architecture. With native 2K image support and stronger semantic adherence, Qwen Image API can better follow complex visual instructions while preserving finer detail, clearer scene structure, and more polished realism in the final output. Keep Qwen Image 2.0 for standard generation and editing, but choose the newer Qwen Image 2.1 API when you need to combine several visual inputs in a flexible editing workflow.
Qwen-Image 2.0 API brings image generation and image editing into a more unified workflow instead of separating them into different model paths. For products that need both creation and refinement, Qwen-Image 2.0 API offers a more consistent way to handle prompt-based generation and follow-up visual edits within the same model workflow.
Qwen Image Edit API supports practical editing tasks such as background changes, object updates, style adjustments, and text edits in a more efficient workflow. For products that need natural-language image editing with quicker response speed, Qwen Image Edit API is designed to handle those tasks on top of a lighter 7B-parameter model architecture.
Alibaba Qwen Image 2.0 API shows strong benchmark performance across both image generation and image editing. In public AI Arena blind testing, Qwen-Image-2.0 is presented as a unified generation-and-editing model with competitive results in both tasks, supporting its positioning as a single model for creative generation and visual refinement.
Qwen-Image-2.0 ranks No. 3 on the AI Arena text-to-image leaderboard with an Elo score of 1029. This result shows that Qwen Image 2.0 API remains highly competitive for prompt-based image generation, especially for workflows that need strong layout control, detailed rendering, and realistic image output.
Qwen-Image-2.0 ranks No. 2 on the AI Arena image edit leaderboard with an Elo score of 1034. This performance shows that Qwen-Image 2.0 API is also highly competitive for image editing tasks, reinforcing its value as a unified model that can support both generation and editing within the same workflow.
Get started with our product in just a few simple steps...
Qwen Image 2.0 API works well for presentation slides, infographics, and other data-heavy visuals that need both readable text and structured composition. It is a strong fit for layouts that combine titles, labels, charts, icons, and supporting visuals in one image.
Qwen-Image-2.0 API is also suitable for travel posters, itinerary visuals, and bilingual guide-style designs. It fits content that needs scenic imagery, multiple text sections, and a clearer visual hierarchy within a single poster-style layout.
For formats that depend on repeated structure and precise placement, Qwen-Image 2.0 API is a practical option for calendars, comics, and other storytelling layouts. It is especially useful when a visual needs multiple panels, organized grids, or several content blocks arranged in a more controlled way.
Qwen Image 2.0 Pro API is useful for poster design, whiteboard-style visuals, and text-rich image editing tasks that involve adding, updating, or refining text inside an image. It fits workflows that need clearer in-image text control, more polished composition, and higher-quality final visuals.
qwen2/text-to-image and qwen2/image-edit run on POST /api/v1/jobs/createTask. Image edit takes one image to edit and offers more aspect ratios than text-to-image; the full parameter list is in the playground and API docs above. Pricing is shown in the pricing card above.
Image-edit prompts are limited to 800 characters, against 5,000 characters for text-to-image. Set the output format to PNG or JPEG explicitly on both models.
The same seed and prompt on the same model version return the same image. Pass a callBackUrl to receive the result when the task finishes, verify each callback with the HMAC signature described in the Kie.ai webhook guide, and query Get Task Details for the authoritative result. New Kie.ai users receive 80 free credits for testing.