Meta Muse Voice Transcribe Brings Real-Time AI Transcription for 70+ Languages with Multi-Speaker Detection

Meta's new AI transcription model can distinguish between multiple speakers and languages in real-time - Engadget

Meta has introduced its first real-time audio model, Muse Voice Transcribe. The model can handle dictation and transcription for more than 20 speakers and can seamlessly handle multiple languages at once, Meta says. Meta CEO Mark Zuckerberg, who recently returned to X after three years of not posting on the platform, shared an example of the model's ability to handle multiple speakers and languages at once. In the video, the transcription is able to automatically distinguish between multiple speakers and switch between languages. It's even able to pick up on "code-switching" and transcribe sentences that use words from multiple languages. "The model decides when to listen. It waits a little longer on hard words and commits faster on easy ones, using adaptive delay to predict each token and increase accuracy," he explained. "It holds up on messy, real audio too -- trained across 70+ languages (with 25 validated at launch), handles mid-sentence code-switching, and manages hour-long sessions with 20+ speakers. Meta's release comes less than a week after Google Gemini 3.5 Transcribe, its own audio model that boasts similar capabilities. But while Google is baking its audio model into Android and (eventually) Chrome, it's not clear if Meta has plans to integrate Muse Voice Transcribe into its flagship services. For now though, people can experience the new mode's capabilities in Meta's recently released Meta AI Mac app. Because the Mac app is able to power voice-enabled features on other apps, Muse Voice Transcribe will now power dictation features on other services. The model is also available to developers within Muse Code and Meta's Model API. It's priced at $3 for 1,000 audio minutes. Additionally, theres a demo version of the Muse Transcribe on Meta's research blog. Muse Voice Transcribe is the latest release from Meta Superintelligence Lab (MSI), which has been churning out new AI models and tools. In the last few weeks, the company has also introduced its first dedicated coding agent, an open-weight model and the aforementioned Meta AI Mac app.

Meta Muse Voice Transcribe Supports 5 Indian Languages, 70+ Globally

* Muse Voice Transcribe supports more than 70 languages * It can distinguish more than 20 speakers * The model can transcribe code-switched conversations Meta has launched Muse Voice Transcribe with native support for five major Indian languages, including Hindi, Tamil, Telugu, Kannada and Malayalam, as part of support for more than 70 languages globally. The model is the first real-time audio perception model developed by Meta Superintelligence Labs. It combines live speech transcription, speaker separation and endpointing in a single system, while also handling code-switching and conversations with more than 20 speakers. Meta Muse Voice Transcribe Brings Real-Time Transcription Muse Voice Transcribe can turn speech into text as someone is speaking, rather than processing the entire recording afterwards, the company explained in a blog post. It can also identify individual speakers and detect when someone starts or stops talking. Transcription, speaker separation and endpointing happen as the audio comes in, without a separate post-processing step. It can distinguish more than 20 speakers in a recording and work with audio longer than an hour, according to Meta. This allows it to handle long conversations without limiting the transcription to shorter clips. The model processes audio in 80-millisecond chunks and determines when it has heard enough information to transcribe each word. It can wait longer for words that are harder to recognise and commit simpler words sooner. Meta says reinforcement learning helps the model decide these delays while keeping transcription errors low. Meta trained the model on more than 70 languages and extensively validated 25 of them for the initial release. Its language support includes five major Indian languages, including Hindi, Tamil, Telugu, Kannada and Malayalam, the company added in a press release. It can also transcribe conversations in which speakers switch between languages, including when the switch happens within a sentence. Users do not have to change the language manually each time someone switches languages. The model supports language, keyword and context biasing, which helps it identify words using information from the audio and the wider conversation. Meta Muse Voice Transcribe Availability and Pricing Meta has made Muse Voice Transcribe available through the Meta Model API, so developers can use the model for speech transcription in their own applications. Meta AI for Mac and Muse Code already use it for dictation. The API is priced at $3 (roughly Rs. 300) per 1,000 audio minutes, which works out to around $0.18 (roughly Rs. 17) per hour.

Meta Launches Muse Voice Transcribe with Multilingual AI Support

Meta introduced Muse Voice Transcribe, a new AI model designed to listen, transcribe, and distinguish between speakers in real time. The model supports streaming transcription, can identify more than 20 speakers in a conversation, and can detect when a person finishes speaking while audio is being captured. Meta says Muse Voice Transcribe can also switch between languages during a conversation. The model can be fine-tuned to recognize specific names, places, and other terms, adding another layer of flexibility for transcription tasks. Meta stated that Muse Voice Transcribe processes incoming audio in 80-millisecond slices. It uses an adaptive delay system to determine when enough information is available to transcribe a word. Easier segments are processed almost instantly, while more complex phrases receive slightly more processing time. The company trained the system using reinforcement learning, balancing transcription accuracy with lower latency. Meta claims this approach helps improve the trade-off between speed and accuracy. The same architecture also handles and detects conversational endpoints. This allows transcription and speaker tracking to operate within a single system. Also Read: According to Meta, Muse Voice Transcribe was trained on over 70 languages, with 25 languages undergoing extensive validation. These include Chinese, French, Hindi, Japanese, Spanish, and Vietnamese. Meta also demonstrated the model switching between English and Mandarin within the same sentence. The company showcased its ability to process long recordings involving multiple speakers without requiring manual cleanup. Meta sees Muse Voice Transcribe as a building block for more personalized AI assistants that can understand accents, interruptions, overlapping speech, and multilingual conversations. The model is rolling out immediately across several Meta products, including voice dictation in Meta AI and the company's Muse Code tool. It is also available through and Meta AI for Mac. Meta's launch comes as real-time speech systems increasingly move beyond basic transcription, with speaker identification, multilingual conversations, and faster response times becoming key parts of AI-powered voice tools.

Meta launches Muse Voice Transcribe with real-time speaker detection and multilingual support

Meta is rolling out the model across Meta AI, Muse Code, its Model API and Meta AI for Mac. Meta has introduced Muse Voice Transcribe, a new artificial intelligence model designed to listen, transcribe and distinguish between speakers in real time. It can handle streaming transcription, identify more than 20 speakers in a conversation and detect when a person has finished speaking while the audio is being captured. Meta says that the model can also switch between languages mid-conversation and fine-tuned to recognise specific names, places and other terms. Built for speed and accuracy Meta stated that the model processes incoming audio in 80-millisecond slices and uses an adaptive delay system to determine when it has enough information to transcribe a word. The easier segments are processed almost instantly, while more complex phrases receive slightly more processing time. The company trained the system using reinforcement learning, balancing transcription accuracy with lower latency. Meta claims this helps the model improve the trade-off between speed and accuracy. The same architecture also handles speaker changes and detects conversational endpoints, allowing transcription and speaker tracking to work within a single system. Supports over 70 languages In a blog post, Meta stated that the model was trained on over 70 languages with 25 undergoing extensive validation. These include Chinese, French, Hindi, Japanese, Spanish and Vietnamese. Meta also showcased the model switching between English and Mandarin within the same sentence. The company also showcased its ability to handle long recordings involving multiple speakers without requiring manual cleanup. Meta sees the technology as a building block for more personalised AI assistants capable of understanding accents, interruptions, overlapping speech and multilingual conversations. Muse Voice Transcribe is rolling out immediately across several Meta products, including voice dictation in Meta AI and the company's Muse Code tool. The model will also be available through Meta's Model API and Meta AI for Mac.