self-hosted/ai
§01·model · /models

Z-Image Turbo

imageactiveApache-2.0

6B step-distilled variant of Tongyi-MAI's Z-Image, running in 8 inference steps at fp16; the siblings are Z-Image-Base (slower, full quality) and Z-Image-Distilled (compressed further). Sixteen cards carry a recipe here from a 16 GB floor, and four of them are Apple machines — the widest Apple reach of any image model in this catalogue. Apache-2.0.

§02·GPUs that run this model
16 total

benchmarked·~ runs via recipe (not benchmarked)· untested·doesn't fit