Qwen Image 3.0 API
Qwen Image 3.0, Alibaba's newest image model, via WaveSpeed — 2K images with strong prompt understanding, and edits from a reference image. 3¢ an image (4¢ for an edit) on Varosity Credits.
$0.03/img on Varosity credits.
no wavespeedaccount needed · provider's list price, no markup · failed renders never billed
Best for
- 2K images for 3¢
- strong prompt understanding
- edits from a reference image
How to get the best result
Qwen Image 3.0, Alibaba's newest image model: 2K images for 3¢ with strong prompt understanding, and edits from a reference image. Pick instead: nano-banana-2.1 for design layouts and in-image text; flux-1-schnell for 1¢ drafts; gemini-3-pro-image for multilingual text and complex layouts.
Prompting
- Describe the subject, the scene, the style and the composition in plain sentences.
- For an edit, pass referenceImageUrl and describe only the change.
Settings on Varosity
- Always 2K; 3¢ an image, 4¢ as an edit with referenceImageUrl.
- aspectRatio 1:1, 16:9 or 9:16; an edit follows the reference image's shape.
Avoid
- Readable text in the image: signs come out as made-up lettering (use nano-banana-2.1 for text).
- Other shapes such as 4:5 or 21:9: REST /v1/images and MCP generate_image refuse them before anything is charged.
Example prompt
A street food stall at night in Taipei, steam rising from bamboo baskets, neon signs reflected on wet pavement, 35mm photo.
From wavespeed.ai, wavespeed.ai, checked against what Varosity sends.
Call it from your code
One Varosity key works for every model.
curl -X POST https://varosity.ai/api/v1/images \
-H "Authorization: Bearer $VAROSITY_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"prompt": "studio product shot of a matte black water bottle, soft rim light",
"modelId": "qwen-image-3.0",
"aspectRatio": "1:1"
}'Or ask your agent
Connect Claude, Cursor, or any MCP client to https://varosity.ai/api/mcp with your key, then say:
"Use Varosity with qwen-image-3.0 to generate an image of …"Agent setup guide
Alternatives to Qwen Image 3.0
- Nano Banana 2 (Gemini 3.1 Flash Image)production-scale quality at flash speed · legible in-image text$0.07/img
- Nano Banana 2.1newest Nano Banana: design, subject consistency, prompt adherence · legible in-image text$0.04/img
- Nano Banana Pro (Gemini 3 Pro Image)best-in-class legible in-image text · multilingual text$0.14/img
- Seedream 5.0 Pronewest Seedream: strong prompt following · edits from a reference image$0.05/img
- Flux Dev12B parameter model · strong prompt adherence$0.02/img
- FLUX.1 [schnell]fast (about 5 s) · good prompt adherence$0.01/img