modal - Run Python workloads on Modal cloud
Deploy and run Python workloads on Modal, including GPU computing, model serving, web endpoints, scheduled jobs, storage, and autoscaling.
Tags
Updated: 2026-09-30serverless computingcloud PythonGPU computingAI/ML deploymentmodel servingweb endpointsscheduled jobsautoscalingpersistent storage
Capabilities
Typical Inputs
Typical Outputs
What this skill does
- Run serverless Python functions
- Deploy AI/ML models
- Run GPU workloads
- Build custom container images
- Serve web endpoints
- Schedule recurring jobs
- Manage persistent volumes
- Pass secrets securely
- Scale containers automatically
- Process inputs in parallel
Inputs
- Python source code
- Model weights
- Datasets
- Modal credentials
- GPU configuration
- Container configuration
- Job schedules
- Persistent storage configuration
Outputs
- Deployed cloud functions
- Running containers
- Web endpoint URLs
- GPU computation results
- Scheduled job runs
- Persistent stored files
- Scaling and execution status
Requirements
- Python 3.10–3.14
- Modal Python SDK
- Modal authentication tokens
- Network access to Modal
- Modal account for cloud execution
