Generate from a prompt
Write a description of up to 4,500 characters, choose Qwen Image 2.0 for speed or 2.0 Pro for finer detail, set the canvas and size, then generate 1, 2, 4, or 6 variations in one request.
How to Use Qwen Image 3.0 and Pro Models.
QwenImage3 is an independent image studio around the Qwen Image family. Sign up for trial credits, generate with Qwen Image 3.0, 3.0 Pro, 2.0, or 2.0 Pro, edit existing images with plain language, and download the results — all in one workspace.
Qwen Image 3.0, 3.0 Pro, 2.0, and 2.0 Pro are available for generation and image editing on QwenImage3. Choose the model that fits your brief and review each result before publishing.
AT A GLANCE
One workspace for every image task.
Generate new images, edit existing ones, and manage the results without leaving the panel.
Edit with plain language
Upload one to three reference images of up to 10 MB each and describe only the change you want. The model preserves the details you ask to keep, so a scene can be restyled without being redrawn.
Download and manage
Save any result at the resolution you chose, from 1K to 2K. Credits are deducted only when a generation succeeds, and your balance and account stay in sync across the site.
QUICK ANSWER
The whole workflow in ten steps.
Sign in, choose a model and mode, describe the image, set the canvas, generate, and refine. The rest of this page explains each step.
- 01Create a free account to receive trial credits.
- 02Choose a model: Qwen Image 3.0 / 3.0 Pro for premium output, 2.0 (fast), or 2.0 Pro (best detail).
- 03Select Text to Image or Image Edit.
- 04Describe the image, or the change you want, in plain language.
- 05Add a negative prompt only when it helps avoid specific issues.
- 06Set the aspect ratio and image size for the final canvas.
- 07Choose how many outputs to create and whether to reuse a seed.
- 08Enable Improve my prompt for short ideas, and add a watermark if needed.
- 09Generate — credits are used only when the request succeeds.
- 10Review the results, download your favorite, and refine one detail at a time.
WORKSPACE OVERVIEW
Understand the controls before you create.
The generator keeps every important setting on one panel: model, mode, prompt, negative prompt, canvas, outputs, seed, and toggles. Results appear in the preview area beside the controls.
Model switcher
Choose Qwen Image 3.0 / 3.0 Pro, 2.0 (fast), or 2.0 Pro (best detail) at the top of the panel.
Text to Image / Image Edit
Switch between creating a new image and editing an uploaded one with two tabs.
Prompt
Describe the subject, environment, lighting, style, camera, and mood in up to 4,500 characters.
Negative prompt
Optional. Name only the problems that matter, such as blurry details, distorted hands, or unreadable text.
Aspect ratio
Choose 1:1, 16:9, 9:16, 4:3, or 3:4 so the composition matches the final destination.
Image size
Pick 1K or 2K output. The panel shows the exact pixel dimensions for the selected ratio.
Outputs
Generate 1, 2, 4, or 6 variations in one request. Each image costs credits.
Random seed
Leave Automatic for fresh variation, or enter a whole number of zero or greater to reproduce a result.
Improve my prompt & watermark
Expand short ideas automatically (on by default) and optionally add a watermark to outputs.
The prompt has more influence than any single setting.
Start with a clear description, generate one image, and improve the prompt gradually before changing several controls.
MODEL GUIDE
Which Qwen Image model should you use?
Match the model to your task before changing prompt settings. These quick guides explain when to use Qwen Image 3.0, 3.0 Pro, 2.0, and 2.0 Pro.
How to use Qwen Image 3.0 for text-to-image generation
Choose Qwen Image 3.0 for long prompts, text-rich posters, infographics, multilingual layouts, realistic scenes, and knowledge-heavy visual briefs. Start with one 1K output while refining composition, then move to 2K for the final image.
Explore Qwen Image 3.0How to use Qwen Image 3.0 Pro for high-quality images
Choose Qwen Image 3.0 Pro when fine detail and higher-fidelity delivery matter more than credit cost. It is suited to polished commercial concepts, detailed product visuals, dense layouts, and final 2K exports.
Explore Qwen Image 3.0 ProHow to use Qwen Image 2.0 for fast AI image generation
Choose Qwen Image 2.0 for quick everyday generation, prompt testing, social graphics, and straightforward image edits. Its balanced speed and credit cost make it useful for exploring several creative directions.
Explore Qwen Image 2.0How to use Qwen Image 2.0 Pro for detailed image editing
Choose Qwen Image 2.0 Pro for finer detail, higher-fidelity edits, and production work that benefits from additional quality. Upload a clear reference image and state exactly what must change and what must remain unchanged.
Explore Qwen Image 2.0 ProTEXT TO IMAGE
How to create an image from a prompt.
Follow these steps in order the first time you use the generator.
Create an account and claim trial credits
Sign in or create an account to unlock the generator. New accounts receive trial credits, and the button shows Start Free Trial until you are signed in. Signing in keeps your credits and account state in sync so a generation can be billed correctly.
Choose a model
The model switcher offers four options. Qwen Image 3.0 and 3.0 Pro are available for generation and editing, with 3.0 Pro targeting higher-resolution, higher-fidelity output. Qwen Image 2.0 is the fast, balanced choice for everyday generation, and 2.0 Pro targets finer detail at a higher credit cost per image.
Select Text to Image
Text to Image creates a brand-new image from your written description. It suits illustrations, product photography, posters, social graphics, character concepts, marketing visuals, wallpapers, and creative artwork. If you already have a source image to change, switch to Image Edit instead.
Write a complete prompt
The prompt field accepts up to 4,500 characters. Describe the main subject, environment, lighting, visual style, camera angle, and overall mood. When a title, label, price, or button copy must appear in the image, put the exact words in quotation marks so they are treated as required text.
Add a negative prompt when needed
The negative prompt is optional. Use it to avoid concrete problems such as blurry details, distorted hands, extra fingers, watermarks, or unreadable text. Keep the list short and relevant to the current image instead of adding every possible defect.
Set the aspect ratio and image size
Aspect ratio choices are 1:1, 16:9, 9:16, 4:3, and 3:4. Image size choices are 1K and 2K, and the panel shows the exact width by height for the selected combination. Square works well for listings and social posts; landscape suits presentations and banners; portrait suits posters and mobile content.
Choose the number of outputs
Outputs controls how many variations are generated in one request: 1, 2, 4, or 6. Each output costs credits, so the button total updates as you change the count. Start with one image while you refine your prompt, then increase the count when the direction is stable.
Set the random seed
The seed field defaults to Automatic, which gives a fresh interpretation of the prompt each time. You can enter a whole number of zero or greater when you want to reproduce or build on an earlier result. For most exploration, Automatic is the simplest choice.
Use Improve my prompt and watermark
Improve my prompt is on by default: it expands short ideas into more descriptive instructions before generation. Turn it off when you already wrote a precise prompt. The watermark option adds a visible watermark to the output when you need it.
Generate, review, and download
Click Generate to start the request. While it runs, the panel shows progress and your results appear with individual download links. Credits are used only when the request succeeds. If the result misses the mark, change one element at a time — lighting, camera, style, or composition — and generate again.
PROMPT STRUCTURE
Give the model a complete creative brief.
A dependable prompt moves from the subject to the mood in a logical order.
01Subject
02Environment
03Lighting
04Style
05Camera
06Mood
COPY-AND-USE TEMPLATE
A [subject] in [environment], illuminated by [lighting], created in [style], captured with [camera], creating a [mood] atmosphere.
A natural editorial portrait of a fictional ceramic artist in a sunlit studio, soft window light, realistic photography, eye-level camera, calm and focused atmosphere.
A clean product photograph of an original portable reading lamp on a small oak desk at dusk, gentle shadows, premium commercial photography, three-quarter angle, warm and minimal atmosphere.
IMAGE EDIT
Change an existing image with plain language.
Image Edit begins with a reference image. Describe what should change and clearly protect the details that should remain the same.
Upload your reference image
Switch to the Image Edit tab and choose one to three reference images. PNG, JPG, and WebP files are supported, up to 10 MB each. Clear, high-resolution sources give the model more detail to preserve.
Describe only the change
The prompt field switches to Describe what you want to change. Focus on the modification: replace the background, change a color, remove an object, or apply a new style. You do not need to redescribe the whole scene unless the transformation is intentional.
Protect what should stay the same
For portraits, products, and brand assets, name what must remain stable: the person's face, expression, clothing, pose, product shape, colors, or lighting. This gives the edit a clear boundary and keeps the identity of the source image intact.
Generate, compare, and refine
Image Edit uses the same credit model as generation and charges only when the request succeeds. Compare the edited result with the original, confirm the requested change was applied, and adjust the instruction gradually if another revision is needed.
BEFORE
A product photo with a plain studio background.
EDITING PROMPT
Replace the studio background with a bright modern office while keeping the product shape, label, colors, and lighting unchanged.
AFTER
The same product and lighting appear in the new environment with the important details preserved.
CREDITS
Understand how credits work.
Generation runs on a transparent credit system: new accounts get trial credits, each image has a clear cost, and credits are charged only when a request succeeds.
Trial credits on sign-up
Creating an account gives you free trial credits, so you can test the full workflow before buying a plan.
Clear per-image cost
Qwen Image 2.0 uses 2 credits per image and Qwen Image 2.0 Pro uses 4 credits per image. The button shows the total for the selected number of outputs before you generate.
Charged only on success
Credits are used only when a generation request succeeds. A failed or blocked request should not consume usable credits.
Plans on the pricing page
Buy more credits anytime on the pricing page. Plans suit occasional experiments, regular content work, and larger production runs.
PROMPT WRITING TIPS
Improve the description before adding complexity.
Better prompts are the fastest way to make results more specific and useful.
Describe the subject clearly
Replace a broad subject such as "a dog" with a specific description such as "a golden retriever sitting beside a mountain lake."
Add an environment
Give the subject a useful context: a modern office, Japanese garden, snowy mountain, luxury apartment, futuristic city, or cozy coffee shop.
Mention the lighting
Use light to shape atmosphere: soft morning light, golden hour, dramatic studio light, overcast daylight, neon light, or candlelight.
Choose a visual style
Name a clear direction such as realistic photography, cinematic, anime, watercolor, oil painting, pixel art, 3D render, or minimalist illustration.
Put exact text in quotes
When a headline, label, price, or button copy must appear in the image, write the exact words inside quotation marks and keep them together.
Refine one change at a time
Change only the light, camera, style, or another single element between generations so you can see which adjustment helped.
COMMON MISTAKES
Fix the workflow instead of guessing.
Most disappointing results become easier to correct when you identify one clear cause.
Writing extremely short prompts
A prompt such as "a city" leaves most decisions undefined. Add the location, time, architecture, action, lighting, and visual treatment.
Forgetting the environment
A subject without context can feel generic or incomplete. Describe where the subject is and how it relates to the background.
Ignoring lighting
Lighting affects realism, depth, materials, and mood. Adding one clear lighting direction is one of the fastest ways to improve a prompt.
Making the negative prompt too long
Do not bury the creative brief under dozens of exclusions. Focus on likely problems that would make this specific image unusable.
Changing too many things at once
A complete rewrite makes it difficult to understand what helped. Keep a strong version and adjust one variable between attempts.
Expecting perfection immediately
AI image creation is iterative. Use each result as feedback, make a focused correction, and compare the new image with the best previous version.
BEFORE YOU GENERATE
Review five things.
- The prompt clearly describes the subject and scene.
- The negative prompt includes only necessary exclusions.
- The aspect ratio matches the intended use.
- Image size and outputs suit the current testing stage.
- Improve my prompt and watermark are set as intended.
How It Works FAQ
Answers about accounts, models, generation, editing, credits, settings, and usage.
Is Qwen Image 3.0 live on this site?
Yes. Qwen Image 3.0 and 3.0 Pro are available on QwenImage3 alongside Qwen Image 2.0 and 2.0 Pro. Choose the model in the generator, review the settings, and generate an image.
Do I need an account to generate images?
Yes. Creating an account unlocks free trial credits and is required before the first generation. Signing in keeps your credits and account state in sync.
How much does a generation cost in credits?
Qwen Image 2.0 uses 2 credits per image and Qwen Image 2.0 Pro uses 4 credits per image. The generate button shows the total for the selected number of outputs before you start.
When are credits charged?
Credits are used only when a request succeeds. A failed or blocked generation should not consume usable credits.
Can I edit an existing image?
Yes. Switch to Image Edit, upload one to three images (PNG, JPG, or WebP, up to 10 MB each), and describe the change in plain language.
Which model should I choose?
Choose Qwen Image 3.0 for long prompts, readable text, multilingual layouts, and knowledge-rich visuals. Choose Qwen Image 3.0 Pro for higher-fidelity final work. Qwen Image 2.0 is the fast, balanced option for everyday generation and editing, while Qwen Image 2.0 Pro adds finer detail at a higher credit cost.
How do I use Qwen Image 3.0 for image editing?
Select Qwen Image 3.0, switch to Image Edit, upload one to three clear reference images, and describe the exact change. State which faces, products, colors, text, composition, or lighting must remain unchanged before generating the edit.
Is Qwen Image 3.0 Pro better for 2K image generation?
Qwen Image 3.0 Pro is the better fit when higher fidelity, small details, dense layouts, and final 2K delivery matter more than generation cost. Test the direction with one output before generating multiple final variations.
Why do the same prompts produce different images?
Generation includes controlled randomness. Leaving the seed on Automatic creates fresh variations; entering a whole-number seed helps reproduce or build on an earlier result when supported.
What settings should I start with?
A clear prompt, a 1:1 ratio, 1K size, and one output is a dependable starting point. Improve my prompt is on by default for short ideas.
Can I use the generated images commercially?
Commercial use depends on the current provider terms, your account, input rights, output content, and applicable law. Review the usage rights page for practical guidance and verify critical text or facts before publishing.
What happens if a generation fails?
A failed or blocked request should not consume usable credits. Fix the prompt or account issue and retry, and contact support if a successful result is not returned.
Start with the model built for the real.
Qwen Image 3.0 and 3.0 Pro are available on QwenImage3 for generation and editing. The full lineup — 3.0, 3.0 Pro, 2.0, and 2.0 Pro — is ready to use today.
Try Qwen Image Free