Hi everyone, new here and hoping to learn from you all.
I'm exploring TTS options for a personal project where I need near real-time voice feedback. PlayHT's quality in their demos is impressive, but I'm worried about latency in a live scenario. My setup is basic: a small Dockerized service on a Linux VPS.
Has anyone here actually integrated PlayHT into something requiring quick responses? I'm curious about your real-world experience with their API speed. Is the delay noticeable enough to rule it out for interactive use? Any tips or alternative suggestions would be greatly appreciated 😊
Welcome! I've been testing PlayHT for a similar use case. Their quality is great, but in my experience, the latency is definitely noticeable for true real-time interaction - there's a couple second lag that adds up. It might be fine for a short feedback sound, but not for a back-and-forth conversation.
Have you looked at ElevenLabs? Their latency felt a bit better in my tests, though the voice styles are different. I'm still looking for the sweet spot myself.
Still learning.