PhoneLLM Released: Sub-100ms Latency Open-Weight Voice Agent Model

Pipecat-ai released PhoneLLM Alpha 1, an open-weight voice model claiming GPT 5.6 Terra-level performance at 1/3 the latency, sub-100ms time-to-first-token, and roughly 1/18 the cost.

2026-08-31 ~ 2026-09-01 · 2 related posts

• PhoneLLM Alpha 1: Matches GPT 5.6 Terra at 1/3 Latency and 1/18 Cost — paf1138 · 2026-08-31
• PhoneLLM: Open-Source Voice Agent Model with Sub-100ms Latency — altryne · 2026-09-01