AIREITER
Z-Image Turbo AI Image Generator and API
Z-Image Turbo is a high-speed text-to-image model built for photorealistic output at a very low cost per image. Write a prompt, pick one of nine aspect ratios, and get a finished image in seconds.
Very low cost per image
Z-Image Turbo is the entry tier for high-volume image generation on AIReiter. Every request costs the same flat rate, which makes budgeting predictable when you generate at scale.
Photorealistic by default
The model is tuned toward realistic lighting, materials, and skin tones rather than stylized illustration. It performs well on product photography, lifestyle scenes, portraits, and any brief where the output needs to read as a photograph.
English and Chinese text rendering
Typography inside the image stays legible in both English and Chinese. That makes the model usable for posters, banners, packaging mockups, and social creative that carries a headline.
Nine aspect ratios, one flat price
Generate at 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 2:1, or 1:2. The ratio does not change the price, so wide banners and tall story frames cost exactly the same as a square.
What to Generate with Z-Image Turbo
High-volume drafting
Generate dozens of candidate images per brief before committing to a slower, more expensive model for the final render. The flat per-image price keeps large exploration batches affordable.
Product and lifestyle photography
Describe the product, setting, and lighting to produce photographic scenes for listings, ads, and catalogue pages. Wide ratios suit banners; square suits marketplace grids.
Social creative with headlines
Because the model renders English and Chinese text cleanly, it can produce social posts and thumbnails that already carry a short headline instead of needing a separate design pass.
Prompt iteration and prototyping
Fast turnaround and low cost make Z-Image Turbo a practical default while you are still tuning a prompt template, testing a pipeline, or demoing a workflow to stakeholders.
Supported Z-Image Turbo Controls
A deliberately small surface. The playground and API expose only what the model actually acts on.
Prompt
Up to 2000 characters, in English or Chinese. Describe the subject, setting, lighting, camera framing, and any text that should appear in the image.
Seed
Optional. Pass an integer to reproduce a similar composition from the same prompt. Omit it to randomize each generation.
Aspect ratios
1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 2:1, and 1:2. The default is 1:1. Custom pixel dimensions are not supported — if you need an exact canvas size, use GPT-Image 2 or Seedream 5.0 Pro on AIReiter.
No resolution tier or batch output
One output tier and exactly one image per request. Pricing is therefore flat, with nothing to compute per call.
Z-Image Turbo Pricing
Flat rate per image
One flat rate per generated image, at any aspect ratio. The exact credit cost is shown in the pricing table above and in the Generate button, and it is the lowest per-image rate among the image models on AIReiter.
One image per request
Each request returns exactly one image. To produce variations, submit multiple requests, optionally with different seeds.
How credits are computed
The submit endpoint returns an estimated credit cost before the task runs. After completion, the query endpoint returns the final credits_used. Failed tasks are not charged.
No hidden surcharges
Because reference images, resolution tiers, and batch output are not supported, and the aspect ratio does not affect the price, the cost is fully determined up front.
Generate with Z-Image Turbo
The same controls apply in the browser playground and through the API. No reference images, no extra setup.
Write the prompt
Describe the subject, setting, lighting, and camera framing. English and Chinese both work. Keep the prompt within 2000 characters.
Choose an aspect ratio
Pick one of nine ratios. Square 1:1 for feeds, 16:9 for banners, 9:16 for vertical stories, 2:1 or 1:2 for panoramic and tall formats.
Optionally set a seed
Pass a seed when you want to reproduce a composition across runs, for example while iterating on wording in the same layout.
Generate and iterate
Submit the task and poll for the result. At this price point, running many prompt variations is cheap — compare outputs, keep the strongest, and move the winner into a higher-tier model if you need a larger final render.
Use Z-Image Turbo through the AIReiter API
Submit one normalized task and query the same external task ID until the generation reaches a terminal state.
Model ID
Send z_image_turbo as the model value in the request body.
Submit a task
POST /api/openapi/submit with the model, params (prompt, aspect_ratio, optional seed), and a unique out_task_id. The response returns a pending status and the estimated credit cost.
Query the result
POST /api/openapi/query with the same out_task_id. Poll until status is completed or failed. Completed responses include the final credits_used and the output image URL.
Asynchronous by design
The submit call returns immediately; use the query endpoint to track progress and retrieve the final image. Download result URLs and store them on your side rather than hot-linking long term.
Related AI Image Models on AIReiter
Krea 2 Turbo
Another fast text-to-image model, tuned for cinematic rather than photographic output, with 1K and 2K tiers and eleven aspect ratios. Choose it when you need higher resolution.
GPT-Image 2
A general-purpose image model with 1K to 4K output, custom pixel sizes, and up to 10 reference images for editing. Choose it when you need image-to-image or multi-reference composition.
Nano Banana Pro
A Google image model with 1K, 2K, and 4K tiers and broad aspect-ratio support, including auto. Choose it when you want reference-guided generation and editing.
Z-Image Turbo FAQ
Answers about generation, supported controls, pricing, and API usage.
/ 01What is Z-Image Turbo?
Z-Image Turbo is a high-speed text-to-image model that produces photorealistic images in roughly eight sampling steps. It renders English and Chinese text cleanly inside the image and is priced as the entry tier for high-volume generation on AIReiter.
/ 02Does Z-Image Turbo support reference images or editing?
No. Z-Image Turbo is text-to-image only. Reference images and image editing are not supported. For image-to-image or multi-reference editing, use GPT-Image 2, Nano Banana Pro, or Seedream 5.0 Pro on AIReiter.
/ 03Which aspect ratios are supported?
Nine presets: 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 2:1, and 1:2, defaulting to 1:1. Custom pixel dimensions are not supported. The ratio you pick does not change the price.
/ 04Can I choose the resolution?
No. The model has a single output tier, so there is no resolution selector and no resolution-based price change.
/ 05How is Z-Image Turbo priced?
A single flat rate per generated image, independent of the aspect ratio you pick. See the pricing table on this page for the current credit cost. Each request returns one image, and failed tasks are not charged.
/ 06Can it render Chinese text in the image?
Yes. The model handles both English and Chinese typography, which makes it suitable for bilingual posters, banners, and social creative that carries a headline.
/ 07Which API model ID should I use?
Use z_image_turbo with POST /api/openapi/submit, then poll POST /api/openapi/query with the same out_task_id until the task is completed or failed.
/ 08Can I reproduce the same image?
Pass the same seed with the same prompt to produce a similar composition. Omit the seed to randomize each generation.