AI 导读
StepAudio 3 ASR 的价格为每小时音频 $0.40，相当于每 1,000 分钟 $6.67，是非流式排行榜上五个最准确模型中最贵的一个。MAI-Transcribe-2 和 Grok Voice Transcribe 2.0 的价格为每 1,000 分钟 $1.67，ElevenLabs Scribe v2 为 $3.67，因此 StepAudio 3 ASR 以更高的每分钟成本换取其在对话音频上的领先优势。
正文
StepAudio 3 ASR is available for $0.40 per hour of audio, equivalent to $6.67 per 1,000 minutes, the most expensive of the five most accurate models on the non-streaming leaderboard. MAI-Transcribe-2 and Grok Voice Transcribe 2.0 cost $1.67 per 1,000 minutes and ElevenLabs Scribe v2 $3.67, so StepAudio 3 ASR trades a higher cost per minute for its lead on conversational audio.
来源：@ArtificialAnlys · x.com