This flux1-schnell model has weights in FP8, which makes running in ComfyUI much faster and use less memory.
Inference Providers
NEW
This model is not currently available via any of the supported third-party Inference Providers, and
HF Inference API was unable to determine this model's library.