Note: Image generation models currently only work on macOS.

Z-Image Turbo is a 6 billion parameter text-to-image model from Alibaba’s Tongyi Lab. It generates high-quality photorealistic images.

Features

• Photorealistic output: Strong at generating realistic photographs, portraits, and scenes
• Bilingual text rendering: Accurately renders both English and Chinese text in images
• Apache 2.0: Open weights available for commercial use

Usage

ollama run x/z-image-turbo "a cat holding a sign that says hello world"

Examples

Simple Prompts

Portrait

A chef in a busy kitchen, steam rising from pots

Landscape

Mountain lake at sunrise, pine trees, morning mist

Product Photos

White sneakers on concrete, overhead shot

Text rendering

A storefront sign that says "BAKERY" in gold letters

Detailed Prompts

Photorealistic portrait

Young woman in a cozy coffee shop, natural window lighting, wearing a cream knit sweater, holding a ceramic mug, soft bokeh background with warm ambient lights, candid moment, shot on 35mm film

Chinese text rendering

Traditional Chinese calligraphy brush painting style, the characters "山高水长" written in elegant black ink on rice paper, red seal stamp in corner, minimalist composition

Creative composition

Surreal double exposure portrait, woman's silhouette filled with blooming cherry blossom trees, soft pink and white petals floating, dreamy ethereal atmosphere, fine art photography

Best Practices

• Use detailed, descriptive prompts (the model excels with rich descriptions)
• For text in images, explicitly specify the text content in quotes and describe style/position
• 1024x1024 resolution is recommended
• The model works well for photorealistic styles out of the box

Limitations

• Complex scenes may occasionally have coherence issues
• The model is not intended to provide factual information
• Output quality is influenced by prompting style

License

This model uses weights from Z-Image-Turbo by Tongyi-MAI (Alibaba), released under the Apache 2.0 license.