Runflow
Alibaba

Z-Image Turbo API

Z-Image Turbo by Alibaba — 6B-param speed model by Tongyi-MAI. 1–8 step inference at $0.005/MP with LoRA and tiling support.

z-image-turbo6B paramsTurbo Diffusion Model
Updated March 2026·Early access
Price
$0.005/MP
Per megapixel · ~$0.005 for 1024×1024
Inference
1–8 steps
Turbo architecture — max 8 steps
Cold Start
5–10s
Not billed, handled by Runflow
Parameters
6B
Turbo Diffusion Model by Tongyi-MAI

Z-Image model family by Alibaba

Multiple Z-Image variants for generation, editing, tiling, and ControlNet — all on the same Runflow API.

Z-Image Turbo

This model

Speed-optimised — 6B params, 1–8 steps, $0.005/MP

Z-Image Turbo (LoRA)

LoRA

LoRA fine-tuning variant for custom styles

Z-Image Turbo Tiling

Tiling

Seamless tiling for textures and patterns

Z-Image Turbo Inpaint

Inpaint

Mask-based inpainting and region editing

Z-Image Turbo ControlNet

ControlNet

Guided generation with structural control

Benchmarked for your use case

Every model on Runflow is scored per vertical — not generic leaderboard numbers. Scores come from domain-expert evaluation on real production prompts.

Use CaseAssessmentNotes
PortraitModerateFast drafts at very low cost — good for previews and iteration
FashionModerateRapid concept generation for high-volume lookbook workflows
ProductModerateQuick product shots for catalogue-scale generation
CreativeStrongGreat for tiling, texture generation, and creative exploration

API Parameters

Full parameter reference for the z-image-turbo model endpoint.

ParameterTypeDescription
promptstringText description of the image to generate.
image_sizeobject | stringOutput size. Presets: square_hd, square, portrait_4_3, portrait_16_9, landscape_4_3, landscape_16_9, auto. Or custom {width, height}. Max 4MP. Default: landscape_4_3.
num_inference_stepsintegerNumber of denoising steps. Range: 1–8. Default: 8. Fewer = faster, more = better detail.
accelerationstringSpeed optimisation level. Options: "none", "regular", "high". Default: "regular".
num_imagesintegerNumber of images to generate per request. Range: 1–4. Default: 1.
output_formatstringOutput format. Options: "jpeg", "png", "webp". Default: "png".
enable_safety_checkerbooleanEnable NSFW content filtering. Default: true.
enable_prompt_expansionbooleanAutomatically expand the prompt for better results. Adds $0.0025 per request. Default: false.
seedintegerRandom seed for reproducible outputs.
sync_modebooleanReturn media as data URI without storing in request history. Default: false.

Technical Specifications

ArchitectureTurbo Diffusion Model by Tongyi-MAI
Parameters6 billion
Inference Steps8 (default) · 1–8 range
Max Resolution4 megapixels
Accelerationnone · regular · high
Output FormatsJPEG · PNG · WebP
LoRA SupportYes — via Z-Image Trainer
ExtensionsInpainting · Tiling · ControlNet · Image-to-Image
LicenseCommercial

Why run Z-Image Turbo on the Runflow API

Flux AI models are available on many platforms. Here's what makes the Runflow image generation API different.

$0.005 per megapixel

One of the cheapest production-grade models available — a 1024×1024 image costs $0.005. Generate thousands of images for the cost of one premium model image.

8-step turbo architecture

Z-Image Turbo compresses inference to 1–8 steps where standard diffusion models need 20–50. Dramatically faster throughput for high-volume pipelines.

Full ecosystem of extensions

LoRA fine-tuning, seamless tiling for textures, ControlNet for structural guidance, inpainting for region editing, and image-to-image — all as API variants.

Acceleration modes

The acceleration parameter (none/regular/high) provides an additional speed layer on top of the already-fast turbo architecture.

Sentinel quality scoring

Every output is scored by Runflow's Sentinel system before delivery, catching low-quality or off-prompt results automatically.

Explore other models

Same API, different strengths. Switch models with one parameter change.

Frequently asked questions

Resources

Start generating images with Z-Image Turbo

Get API access in minutes. No GPU setup, no infrastructure to manage. Pair with Sentinel to control the quality of every AI-generated image you ship.