Hi everyone

[halcyon bloom]

Gemini has been having recurring latency issues on 11labs the last couple of days. Check their status page.

[vagrant shore]

for lower latency while keeping a smart model I really like gpt 5.2 on "none" thinking mode

it averages ~800ms in my case

otherwise on openrouter they seem to have a better stability in terms of uptime

you can add an openrouter llm by selecting the custom model option in elevenlabs