Result data is only kept for 3 days
Guide
Generate human portraits from text with Z-Image AI. Describe the face, expression, wardrobe, lighting, and framing, then choose the available output settings.
Draft campaign concepts, product backdrops, banners, and social graphics from one clear creative brief.
Compare visual directions for characters, environments, layouts, or products before committing to production.
Create supporting illustrations and channel-specific visual variations while keeping the prompt and references consistent.
Not on this page β Z-Image generates from text only. If you need to work from an existing photo, use the AI image editor or an image model that accepts reference images. 1080P, or 2K on a paid plan. The default ratio is 9:16, which suits portrait and social formats; other ratios are available in the same selector.
Z-Image Turbo is tuned for human portraits, including faces, skin, and expression detail. It generates from text only, with no reference image upload.
1080P, or 2K on a paid plan. The default ratio is 9:16, which suits portrait and social formats; other ratios are available in the same selector.
Describe age, build, hair, expression, wardrobe and lighting as separate facts rather than as one adjective pile, and name the shot β headshot, half body, three-quarter. Prompts can run to 2500 characters, so there is room to be specific.
Not on this page β Z-Image generates from text only. If you need to work from an existing photo, use the AI image editor or an image model that accepts reference images.