Cost Comparison: Replicate.ai vs. Hugging Face Inference Endpoint (FLUX.1-schnell Model)
#154
by
rajeshcharhajari
- opened
Hello everyone,
I'm comparing the cost of generating images using the FLUX.1-schnell model on Replicate.ai and Hugging Face Inference Endpoints.
Replicate.ai charges $0.003 per image, meaning $1 can generate approximately 333 images.
Hugging Face Inference Endpoints custom deployment costs $1.80 per hour, with the following configuration:
Nvidia L40s (1 GPU, 48 GB VRAM)
7 CPUs, 30 GB RAM
Has anyone tried deploying the FLUX.1-schnell model on Hugging Face with the given configuration? If so, could you share how many images it can generate in an hour?
Check this screenshot of huggingface deployment config
Thanks in advance for your help!
rajeshcharhajari
changed discussion status to
closed
rajeshcharhajari
changed discussion status to
open