Models · MarkTechPost ·
Microsoft AI Introduces MAI-Transcribe-1.5: 2.4% WER on Artificial Analysis, Best-in-Class FLEURS Accuracy, and Up to 5x Faster Long-Audio Transcription
Microsoft AI released MAI-Transcribe-1.5, its second in-house speech-to-text model. It supports 43 languages, includes keyword biasing for domain terms, is available in Azure AI Foundry, and is reported to transcribe an hour of audio in under 15 seconds.