Alibaba's flagship model surpassed OpenAI on the Artificial Analysis Speech-to-Speech Index with an overall score of 84.1% compared to OpenAI's 79.1%.
OpenAI released GPT-Live-Transcribe and GPT-Transcribe (9.6%–8.98% error); Alibaba's Qwen-Audio-3.0 tops benchmarks at 84.1% vs OpenAI's 79.1%. Voice AI deployment now prioritizes latency vs accuracy; teams evaluate which vendor's tradeoff aligns with production SLAs and transcription ROI.
When you purchase through links on our site, we may earn an affiliate commission. Here’s how it works.
OpenAI launched GPT-Live-Transcribe for low-latency live speech recognition and GPT-Transcribe for asynchronous file processing.
Pradeep Viswanathan
Neowin
@pradeepviswav ·
GPT Transcribe improves on its predecessor but can't catch ElevenLabs, Google, or Mistral on error rates
OpenAI busca revolucionar la transcripción de voz con GPT-Live-Transcribe y GPT-Transcribe, sus nuevos modelos de IA
OpenAI lanza nuevos modelos de voz con IA para transcripción y conversaciones en tiempo real
