From standard language tasks to image processing, our model hub hosts a range of popular models to power your next ML app.
from pipeline import PipelineCloud
api = PipelineCloud("YOUR_API_TOKEN")
# GPT-J Pipeline
model_id = "pipeline_d7502b78863744c495d7d22f321cf7ff"
query = "What is the distance between London and NYC?"
run = api.run_pipeline(model_id, query)
Join over 2,500+ data engineers already using Pipeline Model Hub.
Launch inference or training on cloud GPUs in the blink of an eye. Smart scaling will support your usage as it grows.
Automatically get endpoints for any model or pipeline. You do the data science, our compute engine does the heavy lifting.
No quotas, no minimums, no hourly rental. We charge only for the milliseconds of GPU use your workload needs to complete.
Our team are knowledgeable and friendly. They’ll help you import models and maximise their performance. All within 24 hours, guaranteed.
($0.00000056 per ms)
+ Account access
Custom pipelines for your use-case
24/7 direct support