Litestar Queues lets a Litestar application persist work, run it in a worker, and inspect the result. Use it for work that should outlive the request that started it: sending email, importing files, refreshing reports, or calling a slow service.
Install the package:
pip install litestar-queuesCreate app.py:
from litestar import Litestar, post
from litestar.di import NamedDependency
from litestar_queues import QueueConfig, QueuePlugin, QueueService, task
@task("accounts.sync", queue="accounts", timeout=30)
async def sync_account(account_id: str) -> dict[str, str]:
return {"account_id": account_id, "status": "synced"}
@post("/accounts/{account_id:str}/sync")
async def create_sync_job(
account_id: str,
queue_service: NamedDependency[QueueService],
) -> dict[str, str]:
result = await queue_service.enqueue(sync_account, account_id)
return {"task_id": str(result.id), "status": result.status or "pending"}
app = Litestar(
route_handlers=[create_sync_job],
plugins=[QueuePlugin(config=QueueConfig())],
)Run the application:
LITESTAR_APP=app:app litestar run --reloadEnqueue a task:
curl -X POST http://127.0.0.1:8000/accounts/acct-123/syncThe response contains a task ID and an initial status. QueueConfig() starts
one fresh queue-worker child for this litestar run invocation and shares a
private temporary SQLite file with it. No queue socket or port is exposed, and
the temporary queue is removed on normal shutdown; it is not durable across
server restarts.
Choose where tasks are stored separately from where they run. For durable deployments use a shared backend such as SQLSpec, Advanced Alchemy, Redis, or Valkey. Use standalone workers when the web app and task workers must scale separately. The process-local memory backend remains useful for inline tests or an explicitly single-ASGI-process worker. Cloud Run runs tasks; it does not store them.
- Start here
- Understand the model
- Follow a how-to guide
- Choose backends
- Run an example
- Browse the API reference
Litestar Queues supports Python 3.10 through 3.14 and is licensed under MIT.