v1.0.0์•ˆ์ • ๋ฒ„์ „

Z Image Turbo API

๊ณ ์„ฑ๋Šฅ AI ์ด๋ฏธ์ง€ ์ƒ์„ฑ ๊ธฐ๋Šฅ์„ ์„œ๋น„์Šค์— ํ†ตํ•ฉํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค. ์ด ๋ฌธ์„œ๋Š” ์™„์ „ํ•œ API ๋ ˆํผ๋Ÿฐ์Šค, ์—ฐ๋™ ๊ฐ€์ด๋“œ, ์ฝ”๋“œ ์˜ˆ์ œ๋ฅผ ์ œ๊ณตํ•˜์—ฌ ๊ฐœ๋ฐœ์ž๊ฐ€ ๋น ๋ฅด๊ฒŒ AI ์ด๋ฏธ์ง€ ์• ํ”Œ๋ฆฌ์ผ€์ด์…˜์„ ๊ตฌ์ถ•ํ•  ์ˆ˜ ์žˆ๋„๋ก ๋•์Šต๋‹ˆ๋‹ค.

ํ€ต ์Šคํƒ€ํŠธ

์ƒ์„ฑ ์ž‘์—…์„ ์ œ์ถœํ•˜๊ณ  ์ƒํƒœ๋ฅผ ํด๋งํ•˜๋Š” ๋‘ ๋‹จ๊ณ„๋งŒ์œผ๋กœ ์ฒซ ๋ฒˆ์งธ API ํ˜ธ์ถœ์„ ์™„๋ฃŒํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

1

์ž‘์—… ์ œ์ถœ

POST ์š”์ฒญ์„ generate ์—”๋“œํฌ์ธํŠธ๋กœ ๋ณด๋‚ด Task ID๋ฅผ ๋ฐ›์Šต๋‹ˆ๋‹ค.

2

์ƒํƒœ ์กฐํšŒ

Task ID๋ฅผ ์‚ฌ์šฉํ•ด ์ง„ํ–‰ ์ƒํ™ฉ์„ ํ™•์ธํ•˜๊ณ , ์ด๋ฏธ์ง€ URL์ด ๋ฐ˜ํ™˜๋  ๋•Œ๊นŒ์ง€ ์กฐํšŒํ•ฉ๋‹ˆ๋‹ค.

์ธ์ฆ

HTTP ํ—ค๋”
Authorization: Bearer {YOUR_API_KEY}

์ฃผ์˜: API ํ‚ค๋Š” ์•ˆ์ „ํ•˜๊ฒŒ ๋ณด๊ด€ํ•˜๊ณ , ํด๋ผ์ด์–ธํŠธ ์‚ฌ์ด๋“œ ์ฝ”๋“œ์— ๋…ธ์ถœํ•˜์ง€ ๋งˆ์„ธ์š”.


POST

https://zimageturbo.ai/api/generate

Create a new image generation task. This endpoint is asynchronous and returns a task_id plus initial status; call /api/status to retrieve the final images.

This endpoint responds immediately with task_id. Poll /api/status until the status becomes SUCCESS to obtain image URLs.

๋ณธ๋ฌธ ํŒŒ๋ผ๋ฏธํ„ฐ

ํŒŒ๋ผ๋ฏธํ„ฐ ์ด๋ฆ„ํƒ€์ž…ํ•„์ˆ˜ ์—ฌ๋ถ€์„ค๋ช…
promptstringํ•„์ˆ˜Text prompt for generation (max 1000 characters).
aspect_ratiostringํ•„์ˆ˜Aspect ratio for the generated image. Allowed values: 1:1, 4:3, 3:4, 16:9, 9:16.
{
  "code": 200,
  "message": "success",
  "data": {
    "task_id": "task_1234567890",
    "status": "IN_PROGRESS"
  }
}
GET

https://zimageturbo.ai/api/status

์ž‘์—…์˜ ์‹คํ–‰ ์ƒํƒœ์™€ ๊ฒฐ๊ณผ๋ฅผ ์กฐํšŒํ•ฉ๋‹ˆ๋‹ค.

์ฟผ๋ฆฌ ํŒŒ๋ผ๋ฏธํ„ฐ

ํŒŒ๋ผ๋ฏธํ„ฐ ์ด๋ฆ„ํ•„์ˆ˜ ์—ฌ๋ถ€์„ค๋ช…
task_idํ•„์ˆ˜์ž‘์—…์„ ์ œ์ถœํ•  ๋•Œ ๋ฐ˜ํ™˜๋œ ID์ž…๋‹ˆ๋‹ค.
{
  "code": 200,
  "message": "success",
  "data": {
    "status": "SUCCESS",
    "task_id": "xxxxxxxx",
    "request": {
      "prompt": "xxx",
      "size": "1024*1024",
      "seed": "-1"
    },
    "response": [
      "https://xxx.com/sample.jpeg"
    ],
    "consumed_credits": 15,
    "created_at": "2025-12-05 13:05:09",
    "error_message": null
  }
}

๊ฒฐ์ œ

์š”์ฒญ๋‹น ๋น„์šฉ

$0.02 / ์š”์ฒญ

์‹คํŒจํ•œ ์ž‘์—…์—๋Š” ๋น„์šฉ์ด ์ฒญ๊ตฌ๋˜์ง€ ์•Š์Šต๋‹ˆ๋‹ค.

์˜ค๋ฅ˜ ์ฒ˜๋ฆฌ

  • 401 Unauthorized์œ ํšจํ•˜์ง€ ์•Š์€ API ํ‚ค์ž…๋‹ˆ๋‹ค.
  • 402 Payment Required์ž”์•ก์ด ๋ถ€์กฑํ•ฉ๋‹ˆ๋‹ค.
  • 429 Too Many Requests์š”์ฒญ ํ•œ๋„์— ๋„๋‹ฌํ–ˆ์Šต๋‹ˆ๋‹ค.

Z Image Turbo ์†Œ๊ฐœ

Z Image Turbo๋Š” ์ตœ์‹  ํ™•์‚ฐ(๋””ํ“จ์ „) ๋ชจ๋ธ์„ ๊ธฐ๋ฐ˜์œผ๋กœ ์ตœ์ ํ™”๋œ ์ด๋ฏธ์ง€ ์ƒ์„ฑ ์—”์ง„์ž…๋‹ˆ๋‹ค. ์ถ”๋ก  ์†๋„๋ฅผ ๊ธฐ์กด ๋ชจ๋ธ์˜ 300% ์ˆ˜์ค€๊นŒ์ง€ ๋Œ์–ด์˜ฌ๋ฆฌ๋ฉด์„œ๋„ ๋งค์šฐ ๋†’์€ ํ™”์งˆ์„ ์œ ์ง€ํ•˜๋Š” ๋ฐ ์ดˆ์ ์„ ๋งž์ถ”๊ณ  ์žˆ์Šต๋‹ˆ๋‹ค. ์‹ค์‹œ๊ฐ„ ์ธํ„ฐ๋ž™์…˜, ๊ฒŒ์ž„ ์—์…‹ ์ƒ์„ฑ, ์ด์ปค๋จธ์Šค ๋””์ž์ธ ๋“ฑ์˜ ์‹œ๋‚˜๋ฆฌ์˜ค์— ์ ํ•ฉํ•ฉ๋‹ˆ๋‹ค.

Z-Image-Turbo โ€” 6B-parameter, ultra-fast text-to-image

Z-Image-Turbo is a 6B-parameter text-to-image model from Tongyi-MAI, engineered for production workloads where latency and throughput really matter. It uses only 8 sampling steps to render a full image, achieving sub-second latency on data-center GPUs and running comfortably on many 16 GB VRAM consumer cards.

Ultra-fast generation with production-ready quality

Where many diffusion models need dozens of steps, Z-Image-Turbo is aggressively optimised around an 8-step sampler. That keeps inference extremely fast while still delivering photorealistic images and reliable on-image text, making it a strong fit for interactive products, dashboards, and large-scale backendsโ€”not just offline batch jobs.

Why it looks so good?
  • โ€ข Photorealistic output at speed โ€” Generates high-fidelity, realistic images that work for product photos, hero banners, and UI visuals without multi-second waits.
  • โ€ข Bilingual prompts and text โ€” Understands prompts in English and Chinese, and can render multilingual text directly in the imageโ€”helpful for cross-market campaigns, posters, and screenshots.
  • โ€ข Low-latency, low-step design โ€” Only 8 function evaluations per image deliver extremely low latency, ideal for chatbots, configuration tools, design assistants, and any โ€œclick โ†’ imageโ€ experience.
  • โ€ข Friendly VRAM footprint โ€” Runs well in 16 GB VRAM environments, reducing hardware costs and making local or edge deployments more realistic.
  • โ€ข Scales for bulk generation โ€” Its efficiency makes large jobsโ€”catalogues, continuous feed images, or auto-generated thumbnailsโ€”practical without blowing up compute budgets.
  • โ€ข Reproducible generations โ€” A controllable seed parameter lets you recreate a previous image or generate small, controlled variations for brand safety and experimentation.
How to use
  • prompt โ€“ natural-language description of the scene, style, and any on-image text (English or Chinese).
  • size (width / height) โ€“ choose the output resolution; supports square and rectangular images up to high resolutions (for example, 1536 ร— 1536).
  • seed โ€“ set to -1 for random results, or use a fixed integer to make outputs reproducible.

ยฉ 2025 Z Image Turbo AI Inc. ํŒ๊ถŒ ์†Œ์œ .