kwindla @kwindla · 31 Aug 2026
Everybody building voice agents obsesses about latency. Every enterprise I talk to wants to run LLMs on their own infrastructure (for data management, regulatory/compliance, and cost reasons). PhoneLLM is a new open weights model trained on voice agent tasks: sub-100ms https://t.co/vRsW7ZYhNs
7 586Views
138Likes
10Reposts
8Replies
1Quotes
123Bookmarks
Is that a lot?
1.47×vs this author's median5 148 views is typical
77Percentile for this authorof 13 recent posts
4.93×vs 10K–100K median1 540 views is typical
48.22%Reachviews ÷ followers
2.07%Engagement rateof viewers reacted