Check my GPU →

FLUX.2 [klein] 9B on 6 GB of VRAM

Not entirely. Even Q3_K_M is 1.1 GB over 6 GB, so ComfyUI keeps part of it in system RAM. A distilled 9B FLUX.2 for fast generation and editing in a few steps. Much lighter than FLUX.2 dev.

Image model · 9BAny 6 GB GPUData 2026-10-01
Offload
Not entirely in VRAM, but it can still run.

Nothing fits entirely, but Q3_K_M (4.8 GB) overflows by only about 1.1 GB. ComfyUI keeps that part in system RAM automatically: slower than a full fit, but usable.

Best fileQ3_K_M
File size4.8 GB
VRAM needed~7.1 GB
System RAM32 GB+
Memory map · 6 GB GPU7.1 GB needed · 1.1 GB over 6 GB
04 GB7 GB
Weights Q3_K_M · 4.8 GBWorking memory · 1.5 GBReserve · 0.8 GBSpills to system RAM · 1.1 GB
calculated How this is counted.

Also try the FP8 file (9.4 GB) on NVIDIA GPUs. It does not fit entirely, but a current ComfyUI streams the overflow from RAM and native 8-bit files often beat a small GGUF that does fit. Dynamic VRAM →

01Other amounts of VRAM

8 GB Tight10 GB Runs12 GB Runs well16 GB Runs well20 GB Runs well24 GB Runs well32 GB Runs well

02Every file, on 6 GB

FileSizeNeededOn this cardQualityDownload
16-bit
SAFETENSORS · black-forest-labs
18.2 GB20.5 GB14.5 GB too bigthe original weightsHugging Face →
Q8_0
GGUF · unsloth
10.0 GB12.3 GB6.3 GB too bigpractically identical to the originalHugging Face →
FP8
SAFETENSORS · black-forest-labs
9.4 GB11.7 GB5.7 GB too bigpractically identical to the originalHugging Face →
Q6_K
GGUF · unsloth
7.9 GB10.2 GB4.2 GB too bigvery close to the originalHugging Face →
Q5_K_M
GGUF · unsloth
7.0 GB9.3 GB3.3 GB too bigclose; small differences in fine detailHugging Face →
Q4_K_M
GGUF · unsloth
5.9 GB8.2 GB2.2 GB too biggood; some loss in fine detail and textHugging Face →
Q3_K_M ←
GGUF · unsloth
4.8 GB7.1 GBspills 1.1 GBnoticeable loss of detailHugging Face →
Q2_K
GGUF · unsloth
4.0 GB6.3 GBspills 0.3 GBheavy loss; a last resortHugging Face →

“Needed” adds 1.5 GB of working memory for FLUX.2 klein 9B and a 0.8 GB reserve. Sizes 2026-10-01.

03What to download

PartFileFolderSize
Modelflux-2-klein-9b-Q3_K_M.gguf
Q3_K_M · unsloth/FLUX.2-klein-9B-GGUF
models/unet4.8 GBDownload →
Text encoderqwen_3_8b_fp8mixed.safetensors
Qwen3 8B FP8 · Comfy-Org/flux2-klein-9B
models/text_encoders8.7 GBDownload →
VAEflux2-vae.safetensors
FLUX.2 VAE · Comfy-Org/flux2-dev
models/vae0.3 GBDownload →
Total download · keep about the same free on disk13.8 GB

Alternatives: Qwen3 8B BF16 (16.4 GB). Sizes from Hugging Face, 2026-10-01.

Free ComfyUI workflow, tested on my RTX 5060 Ti: GGUF Q8_0 workflow (.json). Drag the file into ComfyUI. All workflows →

04The text encoder

At 8.7 GB even the smallest Qwen3 8B file is too big for 6 GB, so it runs from system RAM. That slows down reading the prompt, not the picture. Sizes: 16-bit 16.4 GB, FP8 8.7 GB. Why the encoder does not count against VRAM →