Qwen3 Text to Image
Generate an image from a text prompt with Qwen3 Image.
Use Muapi to call Qwen image generation and editing models through one API, including Qwen3 Image, Qwen Image 2.0, the original Qwen Image family, and Qwen 2.1.This page covers Qwen media endpoints available through Muapi. It does not provide Alibaba Cloud's general Qwen chat or language-model API. Check each model page for current parameters, availability, and pricing.
Generate images with the Qwen Image 2.0 Pro endpoint.
Edit images using one or more reference images and a text instruction.
The Muapi Qwen API brings together live Qwen image-generation and image-editing endpoints. Select a model family for text-to-image generation, reference-image edits, multi-image workflows, or supported LoRA operations.
Muapi's Qwen catalog is focused on image media. For the broader endpoint lineup, visit the Qwen Image and Qwen 2.1 family pages. This is not a general Qwen chat or LLM API.
Choose among Qwen3 Image, Qwen Image 2.0, original Qwen Image, and Qwen 2.1 generation endpoints.
Send an instruction prompt with one or more image URLs to supported Qwen image-editing endpoints.
The Qwen Image family includes LoRA-assisted generation and editing plus dataset-based LoRA training endpoints.
| Family | Input | Output | Starting price | Full lineup |
|---|---|---|---|---|
| Qwen3 Image | Prompt or reference images | Generated or edited image | From $0.03/request | /qwen-image |
| Qwen Image 2.0 | Prompt or reference images | Generated or edited image | From $0.04/request | /qwen-image |
| Original Qwen Image | Prompt or reference images | Generated or edited image | From $0.03/request | /qwen-image |
| Qwen 2.1 | Prompt or reference images | Generated or edited image | See current endpoint pricing | /qwen-2.1 |
| Qwen Image LoRA | Prompt, image references, or dataset ZIP | Images or trained LoRA weights | See current endpoint pricing | /qwen-image |
Prices, accepted fields, and availability vary by endpoint. Open the linked family page for current details.
Submit a model-specific request with a Muapi API key, save the returned request ID, and poll the shared prediction result endpoint. Each model has its own accepted request fields.
curl -X POST https://api.muapi.ai/api/v1/qwen3-text-to-image \
-H "x-api-key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"prompt":"A tiny bookshop inside a hollow tree, warm window light, richly detailed illustration"}'
# Response: {"request_id":"REQUEST_ID"}Send a prompt to a Qwen3 text-to-image endpoint.
curl -X POST https://api.muapi.ai/api/v1/qwen3-image-to-image \
-H "x-api-key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"prompt":"Turn the scene into a winter morning while keeping the same composition","images_list":["https://example.com/source.jpg"]}'
# Response: {"request_id":"REQUEST_ID"}Send a prompt and reference image URL list to an image-editing endpoint.
curl -X POST https://api.muapi.ai/api/v1/qwen2-1-text-to-image \
-H "x-api-key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"prompt":"A hand-painted ceramic vase with spring blossoms on a blue table","aspect_ratio":"1:1","resolution":"1K"}'
# Response: {"request_id":"REQUEST_ID"}Choose a supported aspect ratio and resolution for Qwen 2.1 image generation.
curl https://api.muapi.ai/api/v1/predictions/REQUEST_ID/result \ -H "x-api-key: YOUR_API_KEY" # Poll until status is completed, then read the generated image URL from outputs.
Check task status and read the generated image URL from the result.
The live Qwen offering on Muapi includes image generation, image editing, LoRA-assisted image workflows, and Qwen Image LoRA training endpoints.
No. This page covers Qwen image endpoints available through Muapi, not general Qwen chat or text-generation APIs.
No. Required inputs and options vary by endpoint. Check the linked model page and API schema before sending requests.
Create a Muapi API key and send it in the x-api-key header. Keep the request_id from submission and poll the predictions result endpoint until completion.
Use Muapi to access selected Qwen image generation and editing endpoints with one API key and a shared task-result flow.