Elon Musk's Grok Voice 3.0 AI Model Tops Global Voice Agent Rankings
Elon Musk's xAI has unveiled Grok Voice 3.0, an updated AI model capable of "speech-to-speech" interaction with users, bypassing intermediate text conversion. In an independent ranking by Artificial Analysis, the new model scored 86.9 points, surpassing key competitors such as OpenAI’s GPT-Live-1 and Google’s Gemini 3.8 Live.
A major advantage of Grok Voice 3.0 is its high speed. Response latency has been reduced by nearly 40% compared to OpenAI’s solutions, coming in at under one second. The model demonstrates superior speech recognition, maintains conversations confidently, and responds appropriately if interrupted by the user.
The AI assistant is already demonstrating high practical effectiveness in business settings. For instance, in Starlink’s call centers, the system handles approximately 20,000 inquiries daily, independently resolving up to 70% of customer issues without human intervention. The model also achieved record-breaking accuracy during tests involving ServiceNow’s enterprise use cases.
Despite significant performance improvements, the cost of using the technology remains unchanged at $0.08 per minute of conversation (approximately $4.80 per hour). This positions Grok Voice 3.0 as one of the most affordable and fastest solutions on the market for automating support and sales operations.
ORIENT
