在 Big Bench Audio 基准测试中，Phonic 1 (High) 的平均首音频时间（Time to First Audio）为 0.61 秒，比我们语音转语音指数中排名第二的 Grok Voice 3.0（0.81 秒）快约 25%，不到 GPT-Live-1 (Astra, medium)（1.34 秒）和 Gemini 3.8 Live Extended Thinking (High)（1.35 秒）的一半。这使其成为我们语音转语音指数前 5 名模型中最快的一个，其中 Phonic 1 (High)、Grok Voice 3.0 和 Grok Voice 2.0 是前 5 名中仅有的首音频时间低于 1 秒的模型。
Phonic 1 (High)'s average Time to First Audio on the Big Bench Audio benchmark is 0.61 seconds, ~25% faster than Grok Voice 3.0 (0.81s), the next model in our Speech to Speech Index, and less than half the time of GPT-Live-1 (Astra, medium) (1.34s) and Gemini 3.8 Live Extended Thinking (High) (1.35s). That makes it the fastest of the top 5 models in our Speech to Speech Index, with Phonic 1 (High), Grok Voice 3.0 and Grok Voice 2.0 the only models in the top 5 under 1 second.
来源：Artificial Analysis · x.com