ElevenLabs has unveiled Scribe v3 and Scribe v3 Live, its latest speech recognition models designed to deliver accurate transcriptions even amid background noise and crosstalk. According to The Decoder, these new models narrowly outperform those from industry giants such as OpenAI, Google, and Microsoft based on benchmarks from Artificial Analysis.
Despite their accuracy lead, The Decoder also reports that Microsoft and xAI offer speech recognition services that operate significantly faster and cost roughly half as much as ElevenLabs’ Scribe v3. This pricing and speed advantage may influence adoption decisions in competitive markets.
For Japanese investors and tech companies, these advancements highlight the ongoing innovation in AI-driven voice technologies, an area with growing applications across FX trading platforms and automated customer service in Japan’s evolving digital economy.