Microsoft has released MAI-Transcribe-2-Streaming, a real-time speech-to-text model supporting 60 languages with sub-100ms latency and introductory pricing of $0.54 per hour.
Microsoft has released MAI-Transcribe-2-Streaming, a real-time speech-to-text model supporting 60 languages, which leads the industry with a 2.50% word error rate and 0.13-second latency.
A company has built an on-premise speech analytics server to replace cloud-based services, ensuring data sovereignty and improving recognition accuracy for mixed-language telephone calls.
Sarvam AI has introduced Saaras V4, a speech-to-text model supporting 22 Indian languages and global English, featuring a 3-billion parameter architecture and new keyterm prompting capabilities.
SpaceXAI has launched Grok Voice Transcribe 2.0, a speech-to-text model that doubles accuracy and introduces real-time multilingual capabilities for streaming and batch processing.
SpaceXAI released Grok Voice Transcribe 2.0, an API for speech-to-text with improved multilingual accuracy and unchanged pricing starting at $0.10 per hour.
A new open-source macOS dictation tool called Pisary leverages Sber's GigaAM model to provide high-speed, local speech-to-text processing on CPUs.
