wavespeed/FLUX.1-dev-e4m3
Text-to-Image • Updated • 17
Quantized weights we publish
FP8, INT8, NF4 and precision-cast variants of open models. Each card says exactly which tensors changed and which were left alone.
-
-
wavespeed/Qwen-Image-e4m3
Text-to-Image • Updated • 14 -
wavespeed/Qwen-Image-Edit-e4m3
Image-to-Image • Updated • 17 -
wavespeed/Qwen-Image-Edit-l8v1.1-e4m3
Image-to-Image • Updated • 10 • 1 -
wavespeed/Wan2.1-T2V-14B-Diffusers-fp16
Text-to-Video • 14B • Updated • 31 • 1 -
wavespeed/Wan2.1-T2V-14B-Diffusers-fp16-nf4
Text-to-Video • Updated • 18 -
wavespeed/Wan2.1-I2V-14B-480P-Diffusers-fp16
Image-to-Video • 16B • Updated • 22 -
wavespeed/Wan2.1-I2V-14B-720P-Diffusers-fp16
Image-to-Video • 16B • Updated • 310 -
wavespeed/Wan2.1-VACE-14B-bf16
Image-to-Video • 17B • Updated • 222 • 1 -
wavespeed/MAGI-1-24B-distill-quant
Image-to-Video • Updated • 1 -
wavespeed/LongCat-Video-Avatar-1.5-e4m3
Image-to-Video • Updated • 20 -
wavespeed/Ovi-e4m3_e4m3_dynamic_per_tensor
Image-to-Video • Updated • 8 • 1 -
wavespeed/MiniCPM-V-4_5-abliterated-int8
Image-Text-to-Text • 9B • Updated • 20 • 1 -
wavespeed/MiniCPM-V-4_5-abliterated-int4
Image-Text-to-Text • 9B • Updated • 40 • 3