Python SDK
Python SDK
This is the generated API reference for the Hatchet Python SDK. For concepts and guides, see the user guide.
Core pages
- Client: The main Hatchet client, its configuration, and the feature client accessors
- Context: The context object passed to tasks at runtime
- Runnables: Tasks and workflows, the objects you register on workers and run
Feature clients
Feature clients are available as properties on the client, and each covers one area of the Hatchet API. The Guide column links to the user guide page for the feature.
| Client | Description | Guide |
|---|---|---|
| CEL | Evaluate and debug CEL expressions used in event filters and concurrency keys | |
| Crons | Create and manage cron triggers that run workflows on a schedule | Cron Runs |
| Filters | Manage event filters that scope which events trigger runs | Events |
| Logs | Read logs written by task runs | Logging |
| Metrics | Read queue depths, task status counts, and tenant Prometheus metrics | Prometheus Metrics |
| Rate Limits | Create and inspect rate limits that control task throughput | Rate Limits |
| Runs | Get, replay, cancel, and inspect workflow runs and their results | Running Tasks |
| Scheduled Runs | Schedule workflow runs for a specific future time | Scheduled Runs |
| Stubs | Create typed stubs for tasks and workflows registered in another service | Inter-Service Triggering |
| Webhooks | Manage incoming webhooks that trigger runs from external services | Webhooks |
| Workers | List workers and manage their state | Workers |
| Workflows | List and manage registered workflow definitions | Tasks |
Python specifics
- asyncio: Working with asyncio in the Python SDK
- pydantic: Validating task inputs and outputs with Pydantic models
- lifespans: Sharing state like connections and clients across tasks on a worker
- dependency-injection: Injecting dependencies into tasks
- dataclasses: Using dataclasses as task input and output types
Last updated on September 3, 2026