Simple Prompt is instant.
This points to a slowdown in the flow orchestration or state resolution layer.
Has there been a recent deploy or throttling in the flow engine?
Based on the available sources, there was a documented incident on July 14 where Conversation Flow agents experienced delays after function nodes(1). The issue was caused by a message order problem when a function and response occurred simultaneously, which was resolved by 8:30am PDT that day(1).
Additionally, community reports indicate that Conversation Flow agents have experienced issues where nodes don’t carry on as expected(2), and multi-prompt agents sometimes fail to transition between states reliably(3).
If you’re experiencing slowdowns specifically with Conversation Flow orchestration compared to simple prompts, you might consider:
- Checking if your flow has complex tool coordination or multiple decision branches, as the documentation suggests switching to Conversation Flow is beneficial when handling 5+ different functions/tools or multiple decision branches(4)
- Reviewing your node transitions and ensuring proper configuration, as state transition reliability can be improved with explicit transition instructions(3)
I don’t have access to real-time deployment information or current throttling status for the flow engine. For current system status and recent deployments, I recommend:
- Visit Contact Sales
- Ask questions on Discord
- Visit https://github.com/RetellAI/
Would you like a general answer based on my knowledge about typical flow orchestration performance patterns?
**