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

FLUX.1 Fill [dev] on 10 GB of VRAM

Image model · 12BAny 10 GB GPUData 2026-09-25
Runs
Yes, with a compressed file.

Download Q4_K_S (6.8 GB). With the model's working memory it needs about 9.4 GB, leaving 0.6 GB spare on 10 GB. Quality: good; some loss in fine detail and text. For better quality, Q5_K_S (8.3 GB) also runs, with about 0.9 GB spilling into system RAM — a little slower, still practical.

Best fileQ4_K_S
File size6.8 GB
VRAM needed~9.4 GB
System RAM32 GB+
Memory map · 10 GB GPU9.4 GB / 10 GB
05 GB10 GB
Weights Q4_K_S · 6.8 GBWorking memory · 1.8 GBReserve · 0.8 GBFree · 0.6 GB
calculated from real file sizes plus working memory. Not yet measured on this setup. How this works.

01What to download

PartFileFolderSize
Modelflux1-fill-dev-Q4_K_S.gguf
Q4_K_S · YarvixPA/FLUX.1-Fill-dev-GGUF
models/unet6.8 GBDownload →
Text encodert5xxl_fp8_e4m3fn.safetensors
T5-XXL FP8 · comfyanonymous/flux_text_encoders
models/text_encoders4.9 GBDownload →
Text encoderclip_l.safetensors
CLIP-L · comfyanonymous/flux_text_encoders
models/text_encoders0.2 GBDownload →
VAEae.safetensors
FLUX.1 VAE (ae) · Comfy-Org/Lumina_Image_2.0_Repackaged
models/vae0.3 GBDownload →
Total download · keep about the same free on disk12.3 GB

Alternatives: T5-XXL FP16 (9.8 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 12.3 GB, plus roughly 6 GB for Windows, ComfyUI and a browser: 18.3 GB in total. With less RAM it still runs, but Windows starts swapping to disk and loading gets very slow. calculated

02Every file, on 10 GB

FileSizeNeededOn this cardQualityDownload
16-bit
SAFETENSORS · black-forest-labs
23.8 GB26.4 GB16.4 GB too bigthe original weightsHugging Face →
Q8_0
GGUF · YarvixPA
12.7 GB15.3 GB5.3 GB too bigpractically identical to the originalHugging Face →
Q6_K
GGUF · YarvixPA
9.9 GB12.5 GB2.5 GB too bigvery close to the originalHugging Face →
Q5_K_S
GGUF · YarvixPA
8.3 GB10.9 GBspills 0.9 GBclose; small differences in fine detailHugging Face →
Q4_K_S ←
GGUF · YarvixPA
6.8 GB9.4 GBfits · 0.6 GB sparegood; some loss in fine detail and textHugging Face →
Q3_K_S
GGUF · YarvixPA
5.2 GB7.8 GBfits · 2.2 GB sparenoticeable loss of detailHugging 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

T5-XXL: 9.8 GB as 16-bit, 4.9 GB as FP8, 2.9 GB as GGUF Q4_K_M (plus CLIP-L (0.25 GB)). 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 10 GB the FP8 encoder fits on its own, so prompt encoding stays fast.