Datasheet for local AI49 models98 GPUsData read 2026-09-25
No adsNo tracking
Check my GPU →

Can the RTX 4070 Super 12 GB run Krea 2 (Turbo)?

Image model · 12.8B12 GB GDDR6X · 192-bit · 504 GB/s · Ada LovelaceData 2026-09-25
Runs
Yes, with a compressed file.

Download Q5_K_M (8.9 GB). With the model's working memory it needs about 11.5 GB, leaving 0.5 GB spare on 12 GB. Quality: close; small differences in fine detail. For better quality, Q6_K (10.6 GB) also runs, with about 1.2 GB spilling into system RAM — a little slower, still practical.

Best fileQ5_K_M
File size8.9 GB
VRAM needed~11.5 GB
System RAM32 GB+
Memory map · RTX 4070 Super 12 GB11.5 GB / 12 GB
06 GB12 GB
Weights Q5_K_M · 8.9 GBWorking memory · 1.8 GBReserve · 0.8 GBFree · 0.5 GB
calculated from real file sizes plus working memory. Not yet measured on this setup. How this works.

Worth trying on this GPU: with an up-to-date ComfyUI, Dynamic VRAM (on by default for NVIDIA since March 2026) streams whatever does not fit from system RAM, and ComfyUI's own start-up message recommends native FP8/INT8 files over GGUF, saying they “will be faster even if they are larger than your memory”. So before settling for a heavily compressed GGUF, try the native FP8 file (13.1 GB). The file recommended above is the best one that fits entirely — the safe choice on older ComfyUI versions, AMD and Intel. Comfy blog: Dynamic VRAM →

01What to download

PartFileFolderSize
Modelkrea2_turbo-Q5_K_M.gguf
Q5_K_M · vantagewithai/Krea-2-Turbo-GGUF
models/unet8.9 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 disk14.4 GB

Alternatives: Qwen3-VL 4B BF16 (8.9 GB). The 16-bit text encoder is a little more faithful; the smaller one is chosen here because it loads faster and needs less RAM. Sizes read from Hugging Face (2026-09-25). “Download” links start the file directly; the file name links are the same files the official ComfyUI workflows use.

System RAM: 32 GB or more. ComfyUI keeps the model file, the text encoder and the VAE in system RAM and moves them to the GPU as needed. With this set of files that is about 14.4 GB, plus roughly 6 GB for Windows, ComfyUI and a browser: 20.4 GB in total. With less RAM it still runs, but Windows starts swapping to disk and loading gets very slow. calculated

02Every Krea 2 file on 12 GB

FileSizeNeededOn this cardQualityDownload
16-bit
SAFETENSORS · Comfy-Org
26.3 GB28.9 GB16.9 GB too bigthe original weightsHugging Face →
FP8
SAFETENSORS · Comfy-Org
13.1 GB15.7 GB3.7 GB too bigpractically identical to the originalHugging Face →
Q8_0
GGUF · vantagewithai
13.7 GB16.3 GB4.3 GB too bigpractically identical to the originalHugging Face →
INT8
SAFETENSORS · Comfy-Org
13.5 GB16.1 GB4.1 GB too bigpractically identical to the originalHugging Face →
Q6_K
GGUF · vantagewithai
10.6 GB13.2 GBspills 1.2 GBvery close to the originalHugging Face →
Q5_K_M ←
GGUF · vantagewithai
8.9 GB11.5 GBfits · 0.5 GB spareclose; small differences in fine detailHugging Face →
Q4_K_M
GGUF · vantagewithai
7.5 GB10.1 GBfits · 1.9 GB sparegood; some loss in fine detail and textHugging Face →
Q3_K_M
GGUF · vantagewithai
6.0 GB8.6 GBfits · 3.4 GB sparenoticeable loss of detailHugging Face →
Q2_K
GGUF · vantagewithai
4.9 GB7.5 GBfits · 4.5 GB spareheavy loss; a last resortHugging Face →

Sizes from the Hugging Face file listing, read 2026-09-25. “Needed” = file + 1.8 GB working memory for this model + 0.8 GB kept free for the system.

03The text encoder

Qwen3-VL 4B: 8.9 GB as 16-bit, 5.2 GB as FP8. ComfyUI encodes the prompt first and can push the encoder out of VRAM before sampling, so it does not have to fit together with the model. On 12 GB the FP8 encoder fits on its own, so prompt encoding stays fast.

04About this GPU

The RTX 4070 Super 12 GB is an Ada Lovelace GPU with hardware FP8, so ComfyUI can compute Comfy-Org's FP8 files natively: small and fast. (Plain FP8 files use FP8 maths with the --fast fp8_matrix_mult option.)

RTX 4070 Super 12 GB: 12 GB GDDR6X · 192-bit · 504 GB/s · Ada Lovelace. Everything that runs on the RTX 4070 Super 12 GB →

05What more VRAM would change

06Measured and reported results

Nobody has sent measured numbers for this pair yet. If you run Krea 2 on a RTX 4070 Super 12 GB, send your time per image and peak VRAM and it will appear here, credited.

07Questions

How much VRAM does Krea 2 need?

Around 7.5 GB with the smallest file (Q2_K) and 15.7 GB with an 8-bit file (FP8), counting working memory and a small system reserve. The full 16-bit file needs about 28.9 GB.

Which Krea 2 file should I download for the RTX 4070 Super 12 GB?

Q5_K_M (8.9 GB) from vantagewithai/Krea-2-Turbo-GGUF. Load it with Unet Loader (GGUF) from the ComfyUI-GGUF node pack.

Is FP8 faster than GGUF on the RTX 4070 Super 12 GB?

It can be. This GPU has FP8 hardware, and ComfyUI computes FP8 natively for files made for it (Comfy-Org's fp8_scaled files), or for any FP8 file with the --fast fp8_matrix_mult option. GGUF files are unpacked on the fly, which costs some speed.

How much do I need to download for Krea 2 on the RTX 4070 Super 12 GB?

About 14.4 GB for the model file, text encoder and VAE listed on this page, and the same again free on disk. With a 12 GB card, 32 GB of system RAM or more is recommended.