Google unveils Gemini 3.5 speech transcription

Google has announced its new Gemini 3.5 Transcribe model, which significantly improves speech-to-text conversion accuracy for both live audio and previous recordings. The model supports over 85 languages and dialects, and can distinguish up to three speakers. It also understands natural speech, removes unnecessary words, and handles specialized terminology. The model offers two main modes: real-time transcription with latency under one second, and automatic transcription of long recordings. These enhancements greatly boost its speed and accuracy, making it valuable for voice assistants, meetings, and customer service. Currently, it is integrated into products like Gboard on Android and the Gemini app on macOS, with plans to expand to Chrome browser and other platforms.

articleDetails.fontSize.sizeحجم النص

تنويه: هذا ملخص تم إنشاؤه بواسطة الذكاء الاصطناعي

عرض المقال بالكامل