Check my GPU →

Krea 2 (Turbo) on 6 GB of VRAM

Not entirely. Even Q2_K is 1.5 GB over 6 GB, so ComfyUI keeps part of it in system RAM. Krea's own 12B model from June 2026, one of the most downloaded ComfyUI releases of the year. Turbo is the 8-step distilled version; Raw is the base for fine-tuning, same size.

Image model · 12.8BAny 6 GB GPUData 2026-10-01
Offload
Not entirely in VRAM, but it can still run.

Nothing fits entirely, but Q2_K (4.9 GB) overflows by only about 1.5 GB. ComfyUI keeps that part in system RAM automatically: slower than a full fit, but usable.

Best fileQ2_K
File size4.9 GB
VRAM needed~7.5 GB
System RAM32 GB+
Memory map · 6 GB GPU7.5 GB needed · 1.5 GB over 6 GB
04 GB7 GB
Weights Q2_K · 4.9 GBWorking memory · 1.8 GBReserve · 0.8 GBSpills to system RAM · 1.5 GB
calculated How this is counted.

Also try the FP8 file (13.1 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 Tight12 GB Runs16 GB Runs well20 GB Runs well24 GB Runs well32 GB Runs well

02Every file, on 6 GB

FileSizeNeededOn this cardQualityDownload
16-bit
SAFETENSORS · Comfy-Org
26.3 GB28.9 GB22.9 GB too bigthe original weightsHugging Face →
Q8_0
GGUF · vantagewithai
13.7 GB16.3 GB10.3 GB too bigpractically identical to the originalHugging Face →
FP8
SAFETENSORS · Comfy-Org
13.1 GB15.7 GB9.7 GB too bigpractically identical to the originalHugging Face →
INT8
SAFETENSORS · Comfy-Org
13.5 GB16.1 GB10.1 GB too bigpractically identical to the originalHugging Face →
Q6_K
GGUF · vantagewithai
10.6 GB13.2 GB7.2 GB too bigvery close to the originalHugging Face →
Q5_K_M
GGUF · vantagewithai
8.9 GB11.5 GB5.5 GB too bigclose; small differences in fine detailHugging Face →
Q4_K_M
GGUF · vantagewithai
7.5 GB10.1 GB4.1 GB too biggood; some loss in fine detail and textHugging Face →
Q3_K_M
GGUF · vantagewithai
6.0 GB8.6 GB2.6 GB too bignoticeable loss of detailHugging Face →
Q2_K ←
GGUF · vantagewithai
4.9 GB7.5 GBspills 1.5 GBheavy loss; a last resortHugging Face →

“Needed” adds 1.8 GB of working memory for Krea 2 and a 0.8 GB reserve. Sizes 2026-10-01.

03What to download

PartFileFolderSize
Modelkrea2_turbo-Q2_K.gguf
Q2_K · vantagewithai/Krea-2-Turbo-GGUF
models/unet4.9 GBDownload →
Text encoderqwen3vl_4b_fp8_scaled.safetensors
Qwen3-VL 4B FP8 · Comfy-Org/Krea-2
models/text_encoders5.2 GBDownload →
VAEqwen_image_vae.safetensors
Qwen-Image VAE · Comfy-Org/Krea-2
models/vae0.3 GBDownload →
Total download · keep about the same free on disk10.4 GB

Alternatives: Qwen3-VL 4B BF16 (8.9 GB). Sizes from Hugging Face, 2026-10-01.

Free ComfyUI workflow, tested on my RTX 5060 Ti: INT8 workflow (.json). Drag the file into ComfyUI. All workflows →

04The text encoder

At 5.2 GB even the smallest Qwen3-VL 4B 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 8.9 GB, FP8 5.2 GB. Why the encoder does not count against VRAM →