Temporal

Platform

Open-source durable execution engine for long-running workflows

Price: Open source; Cloud credits, then paid
Access: None self-hosted; Cloud API key or mTLS

About

Runs workflows as ordinary code that survives crashes and restarts: state is persisted and failed activities are retried. The MIT-licensed server can be self-hosted, or use Temporal Cloud, billed per action. Official SDKs for eight languages.

What you can do with it

Run multi-step business processes that resume after crashes or deploys
Retry flaky API calls and payments with timeouts and backoff
Keep long-running AI agent loops durable across hours or days

Get started

Install the SDK and the Temporal CLI
Start a local server with temporal server start-dev
Run a worker, then start a workflow from a client

Example

# pip install temporalio; run a local server with: temporal server start-dev
from datetime import timedelta
from temporalio import activity, workflow
@activity.defn
async def greet(name: str) -> str:
    return f"Hello {name}"

@workflow.defn
class SayHelloWorkflow:
    @workflow.run
    async def run(self, name: str) -> str:
        return await workflow.execute_activity(greet, name, schedule_to_close_timeout=timedelta(seconds=10))

Details

Hosting: Hosted service, Self-hosted
Available in: Worldwide
Official SDKs: Go, Java, Python, JavaScript/TypeScript, C#, PHP, Ruby, Rust
MCP server: None