Model reference · open weights
AuraSR is an open-weight image model from fal. AuraSR-v2 (FP32) weighs 1.2 GB; the smallest configuration that runs it is RTX 3060 12 GB.
For self-hosting, compare this build's requirements below with the GPU you plan to rent.
Summary of the fal/AuraSR-v2 model card, 2026-10-01
What it is
| Released by | fal |
|---|---|
| Released | 2024-07-29 |
| Parameters | 618M |
| VRAM | 1.2 GB for the weights |
What it runs on
| Card | One image | Memory |
|---|---|---|
| RTX 3060 12 GB | fits | 11.6 GB |
| RTX 4060 Ti 16 GB | fits | 15.4 GB |
| RTX 3090 24 GB | fits | 23.4 GB |
| RTX 4090 24 GB | fits | 23.4 GB |
| RTX 5090 32 GB | fits | 31.0 GB |
| L40S 48 GB | fits | 44.0 GB |
| A100 80 GB | fits | 78.2 GB |
| H100 80 GB | fits | 78.1 GB |
| RTX PRO 6000 Blackwell 96 GB | fits | 93.8 GB |
| DGX Spark (GB10) 128 GB unified | fits | 107 GB |
| H200 141 GB | fits | 138 GB |
| B200 180 GB | fits | 176 GB |
From the model card
GAN-based Super-Resolution for upscaling generated images, a variation of the GigaGAN paper for image-conditioned upscaling. Torch implementation is based on the unofficial lucidrains/gigagan-pytorch repository.
$ pip install aura-sr
from aura_sr import AuraSR
aura_sr = AuraSR.from_pretrained("fal/AuraSR-v2")
import requests
from io import BytesIO
from PIL import Image
def load_image_from_url(url):
response = requests.get(url)
image_data = BytesIO(response.content)
return Image.open(image_data)
image = load_image_from_url("https://mingukkang.github.io/GigaGAN/static/images/iguana_output.jpg").resize((256, 256))
upscaled_image = aura_sr.upscale_4x_overlapped(image)
Quoted from the model card on Hugging Face. The full card is behind the Hugging Face link above.
How it works
Running it yourself
Rent a machine by the hour. How to run this model is on its model card.