Logan Kilpatrick: experimental feature latency cut from 24+ hours to 10 minutes
TheMoonMidas · x · 2026-10-26
Quoted by TheMoonMidas with a bemused "lol what?", Google's Logan Kilpatrick explained that the current experimental version has 10 minutes of latency (noted with an experimental tag in the screenshot), down from over 24 hours originally.
He added that the team is working on a realtime version and apologized for the hassle, signaling the feature is still early but improving fast.
More from Models
- Ultrafast models unlock a new use case: voice conversations as smart as your chat model — AaronBergman18 · 2026-10-26
- OpenAI and Anthropic Cut LLM Prices by Up to 80% in Months — Why It Happened — minimaxir · 2026-10-26
- Prediction: Google to Overtake OpenAI and Anthropic by Year End — bindureddy · 2026-10-26
- Blogger argues GPT-6 astra crossed his AGI bar: the bottleneck is now us — brandon_galang · 2026-10-26
- Mystery 'Full reset' expiring Nov 30 appears in some Claude accounts — Miyoumu · 2026-10-26
- Frontier models' hidden edge: less scaffolding spend that token-cost models miss — matt_slotnick · 2026-10-26