xAI Grok Voice and OpenAI GPT-Live

xAI's Grok Voice 3.0 scored 86.9 on Artificial Analysis's Speech-to-Speech Index, ahead of GPT-Live-1 at 83.2. On the τ-Voice benchmark, which simulates customer service calls, Grok Voice 3.0 reached a 71.6% task success rate, up from 56.5% under Grok Voice 2.0. Latency came in at roughly 0.8 seconds versus 1.34 seconds for GPT-Live-1, a difference the reporting described as approximately 40% faster. Cost per task on Artificial Analysis's Big Bench Audio run was $0.025 for Grok Voice 3.0, compared to $0.030 for GPT-Live-1 and $0.055 for GPT-Realtime-2.1. The model is deployed in Tesla vehicles handling millions of conversations daily for hands-free tasks. OpenAI then announced GPT-Live-1.1, rolling out to all users of ChatGPT, Codex, and Dots, with API access described as coming soon. The update adds new voices in French, German, Hindi, Mexican Spanish, Korean, and Japanese, improves accents, and reduces filler behavior such as "checking..."