AI models on Craze
Every video, image, voice, music and sound model Craze can direct, with what each one makes and what it costs.
Every video, image, voice, music and sound model Craze can direct, with what each one makes and what it costs.
Starts at$0.004 per image
Alibaba's fast image generator for rapid iteration. Use Z-Image Turbo when speed and affordability matter most.
Z-Image Turbo is built for speed.
Fast inference doesn't mean low quality.
Z-Image Turbo handles readable text well, making it suitable for images that need signs, labels, or typography.
At 6 billion parameters, Z-Image Turbo is architecturally efficient.
Craze creatively jams, researches and plans with you.
Craze generates anything imaginable. Video, audio, music, motion graphics and more.
Craze watches your footage and builds a final cut. With a real editor if you need it.
Every model here runs on every Craze plan, including Free, and Craze can use a different one for each shot.
| Z-Image TurboAlibaba | GPT Image 2.5 SunburstOpenAI | Nano Banana 2.1Google | Nano Banana 2Google | |
|---|---|---|---|---|
| Makes | Text to image | Text + Image to image | Text + Image to image | Text + Image to image |
| Quality | Up to 2K | Up to 1K | Up to 4K | Up to 4K |
| From | $0.004 per image | $0.135 per image | $0.036 per image | $0.071 per image |
| Task | Inputs | Resolution | Aspect ratios | Sound |
|---|---|---|---|---|
| Text to image | Text prompt | 1k, 2k | 21:9, 16:9, 4:3, 3:2, 1:1, 2:3, 3:4, 9:16 | None |
$0/ month
$19/ month
$59/ month
$199/ month
Z-Image Turbo is Alibaba's fast image generator, built for rapid iteration at high efficiency. Created by Alibaba's Tongyi-MAI team, Z-Image Turbo delivers photorealistic results with optimized inference speed and minimal parameters. The model uses only 8 diffusion steps, making it one of the fastest open-weight image generators available while maintaining quality.
Speed and affordability are Z-Image Turbo's defining features. When you need to explore variations quickly or iterate fast within budget constraints, Z-Image Turbo changes what's possible in a single session. The combination of fast generation and low cost makes it ideal for exploratory work where you want to try many directions.
Z-Image Turbo is built for speed. The model uses only 8 diffusion steps and is optimized for efficient generation. That speed means you can explore multiple directions, try variations, and refine without waiting between generations.
Fast inference doesn't mean low quality. Z-Image Turbo maintains photorealistic detail even at maximum speed. You don't trade image quality for generation velocity. This is the key insight: the model achieves speed through architectural efficiency, not by cutting corners on the render itself.
Z-Image Turbo handles readable text well, making it suitable for images that need signs, labels, or typography. It supports bilingual text rendering (English and Chinese), opening possibilities for international work. This is valuable for creators working across markets or building multilingual asset libraries.
At 6 billion parameters, Z-Image Turbo is architecturally efficient. It fits on consumer hardware (16GB VRAM) while delivering results that compete with larger models. Efficiency enables deployment flexibility.
Z-Image Turbo is available under the Apache 2.0 license on GitHub. You can download weights, run the model yourself, or use it through providers. Commercial use is allowed.
Z-Image Turbo is one of the most affordable models on Craze. When you're generating in volume or exploring variations extensively, the cost per image multiplies the value of each session.
Move fast with Z-Image Turbo. You can afford to explore variations because each one returns quickly and costs little.
$0.004 per image is the lowest per-image rate. Resolution and aspect ratio affect the price. Z-Image Turbo is one of the most affordable models on Craze, so you can experiment and iterate without budget constraint. Use the cost calculator on this page to estimate your spend.
When you need to try multiple directions fast, Z-Image Turbo's speed lets you explore in minutes instead of hours. Generate variations of composition, color palette, lighting, and style in a single session.
Generate for daily social posts, quick graphics, and content that moves fast. Z-Image Turbo keeps pace with production speed. Explore variations and publish what works. The speed means you can respond to trending topics or brief changes in direction without losing momentum.
Z-Image Turbo is affordable and fast. Build asset libraries, test variations, and explore creative directions without worrying about cost per generation. Speed plus affordability multiplies exploration.
Generate photorealistic variations of a product, location, or scenario quickly. Try multiple lighting, angles, and compositions in one session. Speed makes variation exploration practical.
Use Z-Image Turbo for exploration, iteration, and speed-critical work. When your priority is trying multiple directions fast and affordably, Z-Image Turbo's efficiency changes what's possible in a session. It excels at volume work where you're generating many variations.
See also: Seedream 5 Lite, Ideogram 4
Describe what you want the way you would brief an editor, and Craze writes the Z-Image Turbo prompts for you. Every Craze plan, including Free, runs Z-Image Turbo and every other media model, and Craze can use a different one for each job. Each image lands in your project, ready to place on the timeline of a video you export as MP4. Generations come out of your plan's monthly credits, from $0.004 per image, with top-ups when you need more.
Rapid exploration and brainstorming: When you need to try multiple directions fast, Z-Image Turbo's speed lets you explore in minutes instead of hours. Social media and fast-moving content: Generate for daily social posts, quick graphics, and content that moves fast. Budget-conscious iteration and variation: Z-Image Turbo is affordable and fast. Photorealistic asset variety: Generate photorealistic variations of a product, location, or scenario quickly.
Z-Image Turbo is Alibaba's fast image generation model from the Tongyi-MAI team, released November 26, 2025. It generates photorealistic images with optimized inference speed.
Alibaba released Z-Image Turbo on November 26, 2025.
Z-Image Turbo is built for speed. According to Alibaba's GitHub, it offers fast inference latency with only 8 diffusion steps. On Craze, that speed makes iteration practical within a single session.
Z-Image Turbo generates at 1K or 2K resolution on Craze. Both run at comparable speeds, so you can choose 2K for sharper detail without significantly longer wait times.
Yes. Z-Image Turbo handles text rendering well, making it suitable for images that need readable text, signs, or typography alongside photorealistic content.
$0.004 per image is the lowest per-image rate. Resolution and aspect ratio affect the price. Z-Image Turbo is one of the most affordable models on Craze. Use the cost calculator on this page to estimate your spend.
every Craze plan, including Free. Every paid plan includes monthly credits, and Z-Image Turbo generations draw from them.
Z-Image Turbo supports ultra-wide 21:9, widescreen 16:9, landscape 4:3 and 3:2, square 1:1, portrait 2:3 and 3:4, and tall 9:16.
Yes. Z-Image Turbo is available under the Apache 2.0 license on GitHub, meaning you can use it for commercial projects and have access to the model weights.
