Z-Image Turbo
A high-speed text-to-image model optimized for fast creative generation, producing detailed, high-contrast, high-resolution images.
The Z-Image API gives you programmatic access to Alibaba's Z-Image model family — Turbo, Base, and P text-to-image generation, plus LoRA support and LoRA training, all through one Muapi API key, from $0.004 per request.
A high-speed text-to-image model optimized for fast creative generation, producing detailed, high-contrast, high-resolution images.
A general-purpose text-to-image model designed for reliable, high-quality image generation from natural-language prompts.
A Z-Image text-to-image variant, the lowest-cost entry point into the family.
Generate images with Z-Image Base while applying a custom-trained LoRA for a specific style or subject.
Generate images with Z-Image Turbo while applying a custom-trained LoRA for a specific style or subject.
Edit an existing image with Z-Image Turbo while applying a custom-trained LoRA.
Train a custom LoRA on Z-Image from a set of reference images, producing reusable model weights.
Train a custom LoRA specifically for Z-Image Base from a set of reference images.
The Z-Image API gives applications programmatic access to Alibaba's Z-Image text-to-image model family. Submit a text prompt and receive a generated image, with a Turbo tier for fast iteration and a Base tier for general-purpose quality.
Muapi exposes the full Z-Image endpoint family through one REST API. Use the image playground to compare Turbo, Base, and P, then train and apply a custom LoRA for a consistent style or subject.
Z-Image Turbo and P generate images at some of the lowest per-request pricing in Muapi's catalog.
Z-Image Base targets reliable, high-quality output for everyday text-to-image jobs.
Apply a custom-trained LoRA to Base or Turbo generations for a consistent style or subject.
Train a reusable LoRA on Z-Image or Z-Image Base from your own reference images.
Use Muapi authentication, billing, and task responses across every Z-Image endpoint.
Submit a task, store its request ID, and poll for the completed image output.
| Endpoint | Input | Output | Price | Best For |
|---|---|---|---|---|
| z-image-turbo | Prompt | Image | $0.007/request | Fast, high-contrast generation |
| z-image-base | Prompt | Image | $0.013/request | General-purpose quality |
| z-image-p | Prompt | Image | $0.004/request | Lowest-cost generation |
| z-image-base-text-to-image-lora | Prompt + LoRA | Image | $0.01/request | Custom style on Base |
| z-image-turbo-text-to-image-lora | Prompt + LoRA | Image | $0.01/request | Custom style on Turbo |
| z-image-turbo-image-to-image-lora | Image + LoRA | Image | $0.01/request | Custom-style editing |
| z-image-lora-trainer | Reference images | LoRA weights | $2.50/request | Training a reusable LoRA |
| z-image-base-lora-trainer | Reference images | LoRA weights | $2.50/request | Training a LoRA for Base |
Prices and supported fields are model-specific. Review the live Playground schema before shipping generated images in production.
Submit a JSON request with your Muapi API key, save the request ID, and poll the shared result endpoint.
curl -X POST https://api.muapi.ai/api/v1/z-image-turbo \
-H "x-api-key: YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"prompt":"A high-contrast studio portrait, dramatic rim lighting","width":1024,"height":1024}'
# Response: {"request_id":"REQUEST_ID"}Call z-image-turbo with a prompt and dimensions.
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.
Poll until the task is completed, then read the returned image URL.
Turbo is optimized for fast, high-contrast generation. Base targets general-purpose, reliable quality. P is the lowest-cost entry point into the family. All three accept a text prompt and return a generated image.
Yes. z-image-base-text-to-image-lora and z-image-turbo-text-to-image-lora apply a custom-trained LoRA during generation, and z-image-turbo-image-to-image-lora applies one during image editing.
Yes. z-image-lora-trainer and z-image-base-lora-trainer train a reusable LoRA from a set of reference images, returning model weights you can reuse across future generations.
Core generation ranges from $0.004 per request (P) to $0.013 per request (Base); LoRA-support generation costs $0.01 per request; LoRA training costs $2.50 per request.
Create a Muapi account, generate an API key, and call the endpoint matching your workflow. Access is available without a separate waitlist.
Create one Muapi API key and connect fast text-to-image generation, LoRA-styled generation, and LoRA training in a single asynchronous workflow.