Skip to content

Inference References

Haoming edited this page Jun 19, 2026 · 8 revisions

Tip

All screenshots below also contain Infotext, which includes the model used, the prompts, and the parameters


SD1


SDXL


Flux.1-Dev


Flux.1-Kontext-Dev


Flux.2-Klein

T2I


Edit


Qwen-Image-Edit

Multi-Image Inputs


Z-Image-Turbo


Anima


Ernie-Image-Turbo


PiD


Wan 2.2

T2I




Tip

For modules (i.e. Text Encoder & VAE), refer to Download Models

Important

When using distilled models (e.g. Flux-Dev, Lightning, Lightx2v), set CFG to 1.0

Caution

Leave Diffusion in Low Bits at Automatic unless you know what you are doing...

Clone this wiki locally