Z-Image AI Image Generator

Photoreal portraits and bilingual text in under a second

Z-Image is no longer offered on TurnVideo. The generator below runs our current image models instead — same workflow, pick any model from its selector.

10 Credits

Recreate This Image

Real prompts and reference shots creators used with Nano Banana models — click Recreate to load one into the generator.

Z-Image AI Image Generator – Instant Photorealistic Images

Z-Image is Tongyi Lab's open-source image model (Apache 2.0), built to be fast and cheap without giving up photorealism — the team positions its 6B-parameter design around sub-second generation. Z-Image is no longer offered on TurnVideo: the generator above runs our current image models instead — Nano Banana Pro, Nano Banana 2 and GPT Image 2 — with the same prompt-in, image-out workflow this page describes.

Last updated August 28, 2026

Z-Image AI Image Generator sample output

Core Capabilities of the Z-Image AI Image Generator

Built for volume drafting

Z-Image was designed to make single images cheap enough to generate in quantity — explore wide, pick ruthlessly, and spend real budget only on the winners. The budget models in the generator above fill that role on TurnVideo today.

Open-source speed, photoreal output

Tongyi Lab released Z-Image under Apache 2.0 with a 6B-parameter architecture designed for sub-second inference. Small and fast does not mean toy-grade here — photorealism is the design goal the team benchmarks it on.

5 aspect ratios

1:1, 3:4, 4:3, 9:16, 16:9 — square posts, portrait stories, widescreen headers. The model generates natively in the framing you pick rather than cropping after.

Text-to-image, honestly scoped

Z-Image generates from a text prompt only — there is no reference-image slot. For prompt-based editing of an existing photo, Nano Banana Pro in the generator above handles reference images.

How Creators Use the Z-Image AI Image Generator

Filling a content calendar at draft prices

Fast, cheap models exist for volume: generate thirty candidates, keep the best five, and save the flagship models for the winners. That drafting workflow is what Z-Image was built for — and it works the same way with the budget models in the generator above.

Photoreal people without a photographer

Photorealistic portraits are the headline strength Tongyi Lab demonstrates for Z-Image — natural skin, believable lighting, everyday scenes rather than uncanny studio gloss. Lifestyle shots for a feed or a lookbook are squarely in its lane.

Posters with English and Chinese text that reads

Bilingual text rendering is one of the model's documented strengths: headlines, labels and layout text in both English and Chinese come out legible, which makes it usable for poster and flyer drafts most models garble.

Scenery and set dressing on demand

Landscapes, cityscapes and establishing shots render with convincing light and atmosphere — useful as blog headers, backgrounds and mood boards where you need a specific place and time of day that stock libraries never quite have.

Getting Better Results from the Z-Image AI Image Generator

Generate from a text prompt

Describe subject, style, lighting and framing in one or two sentences. Concrete beats vague — naming the light and the lens moves the result more than stacking adjectives.

Generate wide, pick ruthlessly

With a fast, cheap model the winning workflow is volume: try five phrasings of the same idea, keep the best, discard the rest. Every result stays in your generation history either way.

Draft cheap, finish on a flagship

Find the composition and mood on a budget model first, then re-run the winning prompt on Nano Banana Pro (50 credits) when the deliverable needs maximum quality or a resolution above 1K.

Leave it running

Generation happens server-side, so closing the tab does not cancel the job — finished images always appear in your generation history.

Z-Image AI Image Generator FAQ

Z-Image is an open-source (Apache 2.0) image generation model from Alibaba's Tongyi Lab, released in late 2025. Its 6B-parameter design targets fast, photorealistic generation with strong bilingual (English and Chinese) text rendering.

No — Z-Image was retired from the TurnVideo catalog in August 2026. The generator on this page runs our current image models (Nano Banana Pro, Nano Banana 2, GPT Image 2) with the same workflow, and images you generated with Z-Image before the retirement stay in your history.

TurnVideo runs on plans rather than free welcome credits: every generation is paid in credits, and each model shows its exact cost next to the Generate button before you run. Failed generations are refunded automatically. The model itself is open source under Apache 2.0, so it can also be self-hosted from Tongyi Lab's official release.

5 aspect ratios (1:1, 3:4, 4:3, 9:16, 16:9) at 1K resolution. For 2K or 4K output on TurnVideo, use Nano Banana Pro or GPT Image 2.

No — Z-Image is text-to-image only. For prompt-based photo editing on TurnVideo, Nano Banana Pro accepts reference images in the generator above.

No. Credits are only consumed for successful generations; failed jobs are refunded automatically.

Yes — images you generate on TurnVideo can be used commercially under your plan terms.