Disty0 / text-to-image updated 4 months ago

FLUX.2-klein-4B-SDNQ-4bit-dynamic

Name: FLUX.2-klein-4B-SDNQ-4bit-dynamic
Author: Disty0

This model uses per layer fine grained quantization. What dtype to use for a layer is selected dynamically by trial and error until the std normalized mse loss is lower than the selected threshold.

Params

2.2 B

Context

—

Downloads 30d

29 K

Likes

Commercial use: allowed apache-2.0 Not gated SAFETENSORS View on Hugging Face ↗

Download history

daily snapshots · 7 days

29 K17 K

Jul 31Aug 2Aug 4Aug 6

Can you run it?

Estimated VRAM at 8K context unless noted. Pick your hardware to see the verdict per quantization.

File	Quant	Size	Est. VRAM	Verdict on RTX 4090 · 24 GB
model.safetensors	u8	5.5 GB	6.8 GB	✅ Runs comfortably

Estimate: file size × 1.1 + KV cache at 8K + 0.5 GB overhead. Not a benchmark — how we calculate this.

Specifications

Parameters: 2.2 B
Tensor type: U8
Licence: apache-2.0
First seen on the Hub: 2026-01-15
Training datasets: undisclosed
Added to our catalog: 2026-07-31

Compare with

stable-diffusion-v1-5 1.6 M stable-diffusion-xl-base-1.0 1.5 M sdxl-turbo 1.1 M Flux2-Klein-9B-True-V2 792 K Qwen-Image-Lightning 438 K GetphatFLUXRealityNSFW 435 K

Appears in

Best open coding models Commercial-use text models Models that fit in 12 GB