Modal

You write Python functions, deploy them to Modal, and get back running compute in the cloud, with no servers for you to manage.

Modal is a serverless platform for AI and data teams that runs workloads at scale, from model inference and batch jobs to scheduled tasks and web endpoints.

Its categories span AI APIs, large language models, image and video generation, OCR and transcription, so the platform is made for teams who run many kinds of models.

Three plans are listed.

Starter costs $0 plus compute each month and includes $30 in free credits, 3 workspace seats, 100 containers and 10 GPU concurrency.

Crons and web endpoints are limited there, though real-time metrics, logs and region selection are included.

Team costs $250 plus compute per month, with $100 in credits, unlimited seats, 1000 containers, 50 GPU concurrency, custom domains, a static IP proxy and deployment rollbacks.

Enterprise uses volume-based pricing and adds audit logs, Okta SSO, HIPAA support, private Slack support and personalized integration help.

The free Starter tier is enough to try Modal before you commit to a paid plan.