Should We Use the Pipedream Python SDK or Directly Call the REST API for Triggering Workflows from Our Python Backend?

This topic was automatically generated from Slack. You can find the original thread here.

Hello everyone :wave:,

We’re deciding between using the Pipedream Python SDK (GitHub - PipedreamHQ/pipedream-sdk-python: Python SDK for Pipedream) or calling the REST API directly.

Our main use case is triggering workflows from our Python backend for our users.

Has anyone run into major limitations with the Python SDK?

Any advice would be much appreciated — thanks so much!
cc

Thanks for your question! Here’s a summary of what you should consider when choosing between the Pipedream Python SDK and calling the REST API directly for triggering workflows from your Python backend: — ## Key Considerations ### 1. Python SDK Convenience - The Pipedream Python SDK is designed to simplify interaction with the Pipedream REST API. It handles authentication, request formatting, and error handling for you. - If your use case is straightforward (e.g., triggering workflows, passing data), the SDK can save you time and reduce boilerplate code. ### 2. REST API Flexibility - Calling the REST API directly gives you full control over all available endpoints and features. - If you need to use endpoints or features not yet supported by the SDK, or if you want to customize requests in ways the SDK doesn’t allow, the REST API is the way to go. ### 3. SDK Limitations - Coverage: The SDK may not support every new or advanced feature of the REST API immediately. If your use case requires the latest or less common API features, check the SDK documentation and issues for coverage. - Updates: The SDK may lag behind the REST API in terms of updates or bug fixes. - Community Feedback: Some users have reported that for very simple use cases (just triggering workflows), the SDK works well. For more complex integrations, direct API calls may be preferable. ### 4. Error Handling & Debugging - The SDK abstracts some error handling, which is convenient, but may make debugging harder if you need low-level details. - Direct API calls let you inspect raw responses and errors, which can help with troubleshooting. — ## Advice - For most standard use cases (triggering workflows, passing data): The Python SDK is a good starting point and will likely meet your needs. - If you hit limitations (missing endpoints, advanced features, or need custom request logic): Switch to calling the REST API directly. - Check the SDK repo for open issues and recent commits to gauge activity and support. — ## Final Tip If you’re just getting started, try the SDK first for its simplicity. If you run into roadblocks, you can always fall back to direct REST API calls—they’re fully supported and documented. If you have specific requirements or run into issues, feel free to share them here or visit Support - Pipedream for more help! — Good luck with your integration!

, can you please guide

I’m not sure but it’s being actively supported by our team

But I believe it’s not 100% ready yet, pending some stuff

understood, thanks :slightly_smiling_face:

so, good to go ahead with python SDK for workflows as kind of superset, and no need to rely on pipedream connect or any client SDKs?

any thoughts?

The Python SDK is not ready yet. It’s usable but the interface to instantiate new clients will likely change soon, so use at your own risk. It does not contain substantial logic though, calling the API directly should be equivalent.

Having said that, we’re very close to releasing the SDK, so if you want to start trying it out you’re more than welcome!

Thanks for your input, As you suggested, we will proceed with the REST API integration to help avoid any potential breaking changes in the near future.

OK! Just FYI, the SDK should be ready within the next week or two

Awesome, appreciate the heads-up