Microsoft AI’s MAI-Transcribe-2 undercuts OpenAI, Google and ElevenLabs on price and speed

Microsoft AI released MAI-Transcribe-2, a speech-recognition model that transcribes audio in 60 languages, adds speaker diarization, word-level timestamps, keyword biasing, automatic language identification, configurable verbatim/clean output, and code-switching for Hinglish and Spanglish, all at $0.10 per hour. Microsoft claims the model achieves a 5.2% average word-error rate, ranking first on the FLEURS benchmark, and reduces enterprise transcription costs from $36,000 to $10,000 for 100,000 hours of audio annually.