modal

Deploy Python workloads using serverless containers that auto-scale from zero to handle demand. You define image-based environments and functions, enabling automatic scaling for ML models, batch processing, and API endpoints.

Yes, GPU-accelerated compute supports machine learning inference and training without infrastructure management. You configure GPU types and resources within serverless containers to execute high-performance tasks.

Configure environments for serverless Python functions using image-based definitions. You enforce configurations for images, resources, secrets, and volumes to establish the exact execution environment required.

The best way to schedule batch processing jobs is using serverless containers with built-in scheduling support. This allows you to run data pipelines and batch tasks across CPU and GPU resources with autoscaling.

Yes, serverless deployment supports persistent storage volumes for data pipelines. You can attach volumes to your containers alongside configured secrets to maintain state and access data across executions.