Hello guys, do you experience latency while using Cartesia TTS and gpt 4.1? It is going over 5 seconds in each iteration

I found OpenAI's LLM to be a cause of significant lag and instability when I tested pipeline agents - and often the requests would time out entirely.

yes iam using the openai 4.1-mini model and i expect some delay in the responses.

no idea tbh, think I've only used/experimented with their

openai/gpt-oss-120b

model - which we moved away from ultimately. Most of our agents use a realtime model - and now we have half-duplex mode working, we can switch the voice to one more suited to our needs.