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

Illustrious XL / Pony (SDXL anime) on 6 GB of VRAM

Image model · 3.5BAny 6 GB GPUData 2026-09-25
Runs well
Yes — and comfortably.

Download Q8_0 (2.7 GB). With the model's working memory it needs about 4.7 GB, leaving 1.3 GB spare on 6 GB. Quality: practically identical to the original.

Best fileQ8_0
File size2.7 GB
VRAM needed~4.7 GB
System RAM16 GB+
Memory map · 6 GB GPU4.7 GB / 6 GB
03 GB6 GB
Weights Q8_0 · 2.7 GBWorking memory · 1.2 GBReserve · 0.8 GBFree · 1.3 GB
calculated from real file sizes plus working memory. Not yet measured on this setup. How this works.

On RTX 40 and RTX 50 GPUs take the FP8 file instead of Q8_0: same class, faster there. On older NVIDIA, AMD and Intel GPUs Q8_0 is the better 8-bit file.

01What to download

PartFileFolderSize
Modelillustrious-q8_0.gguf
Q8_0 · calcuis/illustrious
models/unet2.7 GBDownload →
Total download · keep about the same free on disk2.7 GB

The calcuis FP8 and GGUF files contain the UNet only: take the text encoders and VAE from any SDXL checkpoint, or use the 16-bit checkpoint if it fits. 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: 16 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 2.7 GB, plus roughly 6 GB for Windows, ComfyUI and a browser: 8.7 GB in total. With less RAM it still runs, but Windows starts swapping to disk and loading gets very slow. calculated

02Every file, on 6 GB

FileSizeNeededOn this cardQualityDownload
16-bit
SAFETENSORS · OnomaAIResearch
6.9 GB7.1 GBspills 1.1 GBthe original weightsHugging Face →
Q8_0 ←
GGUF · calcuis
2.7 GB4.7 GBfits · 1.3 GB sparepractically identical to the originalHugging Face →
FP8
SAFETENSORS · calcuis
3.5 GB5.5 GBfits · 0.5 GB sparepractically identical to the originalHugging Face →
Q6_K
GGUF · calcuis
2.1 GB4.1 GBfits · 1.9 GB sparevery close to the originalHugging Face →
Q5_K_M
GGUF · calcuis
1.8 GB3.8 GBfits · 2.2 GB spareclose; small differences in fine detailHugging Face →
Q4_K_M
GGUF · calcuis
1.5 GB3.5 GBfits · 2.5 GB sparegood; some loss in fine detail and textHugging Face →
Q3_K_M
GGUF · calcuis
1.2 GB3.2 GBfits · 2.8 GB sparenoticeable loss of detailHugging Face →
Q2_K
GGUF · calcuis
0.9 GB2.9 GBfits · 3.1 GB spareheavy loss; a last resortHugging Face →

Sizes from the Hugging Face file listing, read 2026-09-25. “Needed” = file + 1.2 GB working memory for this model + 0.8 GB kept free for the system. The 6.94 GB checkpoint holds the UNet plus text encoders and VAE; only the UNet (about 5.1 GB) sits in VRAM while sampling. The GGUF and FP8 files from calcuis contain the UNet only.

03The text encoder

The text encoder is inside the checkpoint, so there is nothing extra to download.