How would I approach handling agent state/instances in a scaleable manner for an HTTP API?

[wintermute-cell]

Hey!

I hope someone can give me a lead here. I'm a bit stumped on how I'd properly approach the situation where I build an agent chat behind an API in such a way that it can be replicated.
The naive solution would be to create a new agent for every http request, e.g. every user message to retain full statelessness. But creating an agent object can be relatively expensive, depending on how the agent is set up.

Any help is appreciated!

[manthanguptaa]

Hey @wintermute-cell! Agno's Agents initialize extremely fast, so overhead in your application will be minimal. To optimize your implementation, define the agent in a variable outside the API function and then call that variable within the function whenever you need to use the agent. Let me know if this helps you