Microsoft launches MAI-Transcribe-2, cutting transcription costs by 72%

Microsoft AI has launched MAI-Transcribe-2, a new speech-recognition model that the company claims outperforms competitors like OpenAI, Google, and ElevenLabs in terms of speed, accuracy, and cost. This model was unveiled on Thursday and is priced at just 10 cents per hour of audio, a dramatic reduction from the previous version, which cost 36 cents per hour. This price cut of approximately 72% is poised to make transcription services more accessible for businesses.

The release of MAI-Transcribe-2 comes as Microsoft accelerates its strategy to develop advanced AI models in-house. Over the past two years, the tech giant has focused on building its capabilities across various modalities, gradually phasing out reliance on OpenAI’s technology. The speed at which Microsoft has rolled out its transcription technology highlights its commitment to becoming a leader in the AI space, particularly in applications that can directly benefit enterprises.

The financial implications for businesses using MAI-Transcribe-2 are significant. For instance, a company processing 100,000 hours of call-center audio annually will see its costs plummet from $36,000 to just $10,000. This level of savings means that transcription becomes a manageable expense, rather than a budgetary concern, which could lead to wider adoption of such technologies across industries.

Looking ahead, Microsoft's aggressive pricing strategy may force competitors to reevaluate their offerings and pricing models. As the demand for efficient transcription services grows, companies may need to innovate to keep pace with the new standard set by MAI-Transcribe-2. This could reshape the competitive landscape in AI-driven transcription, making it a critical area to watch in the coming months.