DogeDesigner· @cb_doge · X·· 3 小时前AI 评分47
AI 导读
突发：在 Artificial Analysis 的 Big Bench Audio 对比中，Grok Voice 3.0 是这些模型里性价比最高的： • Grok Voice 3.0：$0.025/task • GPT-Live-1 (Astra, Medium)：$0.030/task • GPT-Realtime-2.1 (High)：$0.055/task 比 GPT-Realtime-2.1 便宜 50% 以上
正文
BREAKING: Grok Voice 3.0 is the most cost-effective among these models in Artificial Analysis' Big Bench Audio comparison:
- Grok Voice 3.0: $0.025/task
- GPT-Live-1 (Astra, Medium): $0.030/task
- GPT-Realtime-2.1 (High): $0.055/task
Over 50% cheaper than GPT-Realtime-2.1
来源：DogeDesigner · x.com