Face-to-Sticker by fofr logo

Face-to-Sticker by fofr

Paid

Convert faces into sticker-style images with an open-source AI model

5.0
Type
Saas
LinksX

About Face-to-Sticker by fofr

Face-to-Sticker by fofr is an open-source AI model hosted on Replicate that converts facial images into sticker-style graphics. It leverages InsightFace for face detection and supports adding LoRAs for style variation. The model runs on Nvidia L40S GPU hardware, with typical inference completing in about 23 seconds at a cost of approximately $0.022 per run. It is open source (weights are non-commercial only) and can be executed locally via Docker. Created by fofr and using LoRAs by artificialguybr.

Key Features

Converts faces into sticker-style images
Open-source model (weights non-commercial)
Can be run locally with Docker
Uses InsightFace for face detection
Supports LoRAs for additional styles

Pros & Cons

Pros
  • Low cost (~$0.022 per run on Replicate)
  • Open source and can be run locally
  • Fast inference (~23 seconds typical)
  • Supports LoRAs for style customization
Cons
  • Non-commercial use only
  • Output quality may vary based on input face quality
  • Requires Replicate account or local setup for full usage

Alternatives to Face-to-Sticker by fofr

FAQ

How much does it cost to run Face-to-Sticker on Replicate?
Approximately $0.022 per run, though this may vary based on inputs.
Is Face-to-Sticker open source?
Yes, the model is open source and can be run locally with Docker, but the weights are non-commercial only.
How long does a typical prediction take?
Predictions typically complete within 23 seconds, though time may vary based on inputs.
Can I run Face-to-Sticker on my own computer?
Yes, you can run it locally using Docker.
What hardware does it use on Replicate?
It runs on Nvidia L40S GPU hardware.