ThinkingNews
Top Stories
3 outlets·4 reports·over 21h

Meta prices Muse Voice Transcribe at $0.18 an hour, with real-time diarization for 20+ speakers: a steal for enterprises?

ThinkingNews Desk · how this was written

Microsoft AI has launched the MAI-Transcribe-2 speech-recognition model, which it says outperforms competing services such as Gemini 3.5 Transcribe and GPT-Transcribe on benchmark tests. The model transcribes audio in dozens of languages, provides speaker diarization, word-level timestamps, keyword biasing, automatic language identification and code-switching, and is priced at $0.10 per audio hour, a rate locked in through 2026.

Written from all 3 reports below, not from any single one.

How it was reported

  1. VentureBeat·
    Meta prices Muse Voice Transcribe at $0.18 an hour, with real-time diarization for 20+ speakers: a steal for enterprises?

    Meta’s Muse Voice Transcribe streams transcription, endpoint detection, and speaker diarization for up to 20 participants at $0.18 per audio hour, handling hour-long clips and seamless multilingual code-switching. Trained on 70+ languages (25 fully validated), it processes 80 ms audio chunks with adaptive delay and reinforcement-learning rewards to balance word-error-rate and latency, exposing diarization as a first-class mode with turn-level timestamps and session-wide speaker labels.

  2. VentureBeat·
    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.

  3. TechCrunch·
    Meta is paying to peek at how you use their latest AI model

    Meta’s Muse Spark model introduces a “contributor” tier that cuts input token pricing from $1.25 to $0.10 per million and output token pricing from $4.25 to $0.20 per million for users who share prompts and outputs, a roughly 95% discount aimed at collecting data to improve agentic tools. The pricing shift follows Meta’s pivot after ending an employee-tracking project and aligns with similar token-price reductions by Anthropic and OpenAI to remain competitive.

  4. TechMeme·
    Microsoft AI debuts MAI-Transcribe-2, a speech recognition model that it says beats Gemini 3.5 Transcribe and GPT-Transcribe, at $0.10/audio hour through 2026 (Michael Nuñez/VentureBeat)

    Microsoft introduced MAI-Transcribe-2, a speech-recognition model that outperforms Gemini 3.5 Transcribe and GPT-Transcribe on benchmark tests. Microsoft priced the service at $0.10 per audio hour, with the rate locked in through 2026.

Related stories

Share: