Executing Components Locally and on the Server in LangFlow Workflows

[flefevre]

Hello LangFlow community,

I am working in an enterprise environment where specific constraints limit the execution of certain components on the server. Specifically, internet search or connectivity is not allowed on the server due to concerns about potential malicious data. However, users are allowed to perform internet searches directly from their local machines.

I would like to know if LangFlow supports a setup where some components of a workflow can execute locally on the user’s machine while others run on the server (e.g., for GPU processing). For example, if a workflow includes an agent or a component that requires an internet search, could that step be executed locally on the user’s computer instead of the server?

Ideally, I envision a way to export workflows such that certain components or agents are automatically set up as local clients on the user’s computer (Windows, Linux, etc.). This executable would handle tasks locally (like internet searches) and then communicate with other distributed components that require server-side resources, such as GPUs.

Does LangFlow have any mechanisms to achieve this, or would this require custom development? Any guidance or suggestions would be greatly appreciated.

Thank you!