openmodeldb 1.0.0

Browse and download AI upscaling models from OpenModelDB

OpenModelDB

Browse and download AI upscaling models from OpenModelDB.

Install

pip install openmodeldb

CLI

openmodeldb

Select scale → pick a model → download.

Python API

from openmodeldb import OpenModelDB

db = OpenModelDB()
# <OpenModelDB: 658 models>

# List models (formatted table)
db.list(scale=4)
db.list(scale=1, architecture="compact")

# Find models (returns list[Model])
models = db.find(scale=4)
compacts = db.find(scale=1, architecture="compact")

# Search by name, author, tags or description
results = db.search("denoise")

# Download by name or Model object
db.download("4xNomos8k_atd_jpg")
db.download(models[0])
db.download(models[0], dest="./my_models/")

# Download a specific format (pth, safetensors, onnx)
db.download("4xNomos8k_atd_jpg", format="safetensors")

# Download all available formats
db.download_all("4xNomos8k_atd_jpg")

# Silent mode (no output, for use as a library)
path = db.download("4xNomos8k_atd_jpg", quiet=True)

# Get download URL (for custom download logic)
url = db.get_url("4xNomos8k_atd_jpg")
url = db.get_url("4xNomos8k_atd_jpg", format="safetensors")

# Dict-style access
model = db["4xNomos8k_atd_jpg"]
print(model.name, model.author, model.scale, model.architecture)

# Check if a model exists
"4xNomos8k" in db  # True

# Browse architectures and tags
db.architectures()  # ['atd', 'compact', 'cugan', 'dat', ...]
db.tags()           # ['anime', 'denoise', 'photo', ...]

# Iterate
for model in db:
    print(model)

# Launch interactive CLI
db.interactive()

Dependencies

• InquirerPy — interactive prompts
• rich — progress bars and tables
• pycryptodome — Mega.nz decryption

Credits

• OpenModelDB — the open model database
• All model authors and contributors