AI infrastructure that developers love
Run inference and compute in Python without managing servers: pay per second, scale to zero.
Serverless AI infrastructure: containers with per-second billing, GPUs attached in one decorator, scale-to-zero. The Python-first compute layer behind half the AI demo-to-production stories.
Who it's for
ML and AI engineers who want to run Python workloads on autoscaling GPUs with no infra code.
Official links
Keep exploring
Concepts in the glossary
1 in catalog
Modal works with
1 in catalog