Skip to content
Docs menu

Image models

Qwen Image

On this page

Qwen Image — fast, high-fidelity AI image generation & editing.

modelIdqwen-text-to-image, qwen-image-to-image
Modalityimage
PricingSee this model on the Pricing page for the current per-call price (with your markup).

Operations

OperationmodelIdEndpointRequired input
Text To Imageqwen-text-to-imagePOST /api/v1/generateprompt
Image To Imageqwen-image-to-imagePOST /api/v1/generateprompt, image_url
Poll taskGET /api/v1/task/{id}?model=qwen-text-to-image

Text To Image — qwen-text-to-image

FieldTypeRequiredValues / example
promptstringYesThe prompt to generate the image with Max 5000 characters. (default: A terraced rice field at dawn wrapped in fog, farmers walking the ridgeline, soft pastel light, wide landscape.)
image_sizestringNoThe size of the generated image (options: square | square_hd | portrait_4_3 | portrait_16_9 | landscape_4_3 | landscape_16_9) (default: square_hd)
num_inference_stepsnumberNoThe number of inference steps to perform (range 2-250) (default: 30)
seednumberNoThe same seed and the same prompt given to the same version of the model will output the same image every time
guidance_scalenumberNoThe CFG (Classifier Free Guidance) scale is a measure of how close you want the model to stick to your prompt when looking for a related image to show you (range 0-20) (default: 2.5)
enable_safety_checkerbooleanNoThe safety checker is always enabled in Playground. It can only be disabled by setting false through the API. (true/false) (default: true)
output_formatstringNoThe format of the generated image (options: png | jpeg) (default: png)
negative_promptstringNoThe negative prompt for the generation Max 500 characters. (default: )
accelerationstringNoAcceleration level for image generation. Options: 'none', 'regular', 'high'. Higher acceleration increases speed. 'regular' balances speed and quality. 'high' is recommended for images without text (options: none | regular | high) (default: none)
nsfw_checkerbooleanNoA configurable parameter. Defaults to true in the Playground. (true/false) (default: true)
bash
curl -X POST https://you.bot/api/v1/generate \
  -H "Authorization: Bearer $YOUBOT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"modelId":"qwen-text-to-image","input":{"prompt":"A photorealistic portrait of a red fox in a snowy forest, soft morning light, shallow depth of field, 85mm.","image_size":"square_hd","num_inference_steps":30,"seed":1,"guidance_scale":2.5,"enable_safety_checker":true,"output_format":"png","negative_prompt":"blurry, low quality, distorted, watermark, text","acceleration":"none","nsfw_checker":true}}'

Image To Image — qwen-image-to-image

FieldTypeRequiredValues / example
promptstringYesThe prompt to generate the image with Max 5000 characters. (default: Turn this plain teacup photo into a delicate watercolor still-life, softening the porcelain into loose washes of blue and letting the rising steam bleed gently into the paper grain.)
image_urlstringYesThe reference image to guide the generation (image URL)
strengthnumberNoDenoising strength. 1.0 = fully remake; 0.0 = preserve original (range 0-1) (default: 0.8)
output_formatstringNoThe format of the generated image (options: png | jpeg) (default: png)
accelerationstringNoAcceleration level for image generation. Options: 'none', 'regular', 'high'. Higher acceleration increases speed. 'regular' balances speed and quality. 'high' is recommended for images without text (options: none | regular | high) (default: none)
negative_promptstringNoThe negative prompt for the generation Max 500 characters. (default: blurry, ugly)
seednumberNoThe same seed and the same prompt given to the same version of the model will output the same image every time
num_inference_stepsnumberNoThe number of inference steps to perform (range 2-250) (default: 30)
guidance_scalenumberNoThe CFG (Classifier Free Guidance) scale is a measure of how close you want the model to stick to your prompt when looking for a related image to show you (range 0-20) (default: 2.5)
enable_safety_checkerbooleanNoThe safety checker is always enabled in Playground. It can only be disabled by setting false through the API. (true/false) (default: true)
nsfw_checkerbooleanNoA configurable parameter. Defaults to true in the Playground. (true/false) (default: true)
bash
curl -X POST https://you.bot/api/v1/generate \
  -H "Authorization: Bearer $YOUBOT_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"modelId":"qwen-image-to-image","input":{"prompt":"A photorealistic portrait of a red fox in a snowy forest, soft morning light, shallow depth of field, 85mm.","image_url":"https://example.com/input.jpg","strength":0.8,"output_format":"png","acceleration":"none","negative_prompt":"blurry, low quality, distorted, watermark, text","seed":1,"num_inference_steps":30,"guidance_scale":2.5,"enable_safety_checker":true,"nsfw_checker":true}}'