Microsoft Launches MAI-Transcribe-2 for Speech-to-Text at the World's Lowest Price

Microsoft announced the launch of its new specialized speech-to-text model, MAI-Transcribe-2, claiming it is the fastest, most accurate, and cheapest in the world, available via the Azure AI Foundry platform for developers and businesses.

The model features audio transcription services at a very low cost of just 10 cents, and Microsoft confirmed that MAI-Transcribe-2 outperforms OpenAI, Google, and ElevenLabs across speed, accuracy, and cost metrics simultaneously.

This launch came as part of a wave of new artificial intelligence models announced by Microsoft on the Foundry platform, including text, image, and audio models, as part of the company's strategy to build an integrated AI services ecosystem.

MAI-Transcribe-2 represents a step in the escalating competition within the specialized models market, as Microsoft seeks to offer cheaper and faster alternatives to widespread audio transcription models. Demand for speech-to-text solutions is rapidly expanding in business, healthcare, and education applications.

What Do These Terms Mean?

Speech Transcription: Converting spoken speech into written text using artificial intelligence, used in transcribing meetings, voice messages, distance learning, and healthcare.

Azure AI Foundry: Microsoft's cloud platform for artificial intelligence that provides developers and businesses with AI models and tools ready for integration into their own applications.

Specialized Model: An AI model trained on a single specific task such as speech transcription, unlike general models like ChatGPT that perform multiple tasks.