BigLoveKleinFp8

FP8-quantized version of FLUX.2-klein-base-9B by Black Forest Labs.

Model Details

• Base Model: black-forest-labs/FLUX.2-klein-base-9B
• Quantization: FP8 (E4M3FN)
• Architecture: Rectified Flow Transformer (9B parameters)
• License: cc-by-nc-4.0

Description

This is an FP8-quantized variant of the FLUX.2 Klein Base 9B model, optimized for reduced VRAM usage while maintaining high image quality. The FP8 format significantly reduces memory footprint compared to the full-precision model.

Usage

ComfyUI

Place the model file in your ComfyUI/models/diffusion_models/ (or unet) folder and select it in the appropriate loader node.

Diffusers

from diffusers import FluxPipeline
import torch

pipe = FluxPipeline.from_pretrained(
    "Granddyser/BigLoveKleinFp8",
    torch_dtype=torch.bfloat16
)
pipe.to("cuda")

image = pipe(
    prompt="your prompt here",
    num_inference_steps=4,
    guidance_scale=0.0,
).images[0]

image.save("output.png")

License

FLUX.2-klein-base-9B Model is licensed by Black Forest Labs. Inc. under the FLUX.2-klein-base-9B Non-Commercial License. Copyright Black Forest Labs. Inc.