Meta’s Muse Voice Transcribe supports 70+ languages along with 5 Indian languages

Artificial intelligence is making voice transcription faster and more accurate. On September 1^st 2026, Meta announced a new speech to text model called Muse Voice Transcribe. It can transcribe conversations in real time as people are speaking. The model was developed by Meta Superintelligence Labs and is designed to handle code switching, where people switch between two or more languages while talking. It does not need separate models or extra processing to do this. Muse Voice Transcribe can also identify different speakers in a conversation and know when a person has finished speaking. Meta says the model can separate more than 20 different voices in a recording and can also handle audio that is more than an hour long. The model supports more than 70 languages around the world. It also has native support for five major Indian languages which is Hindi, Tamil, Telugu, Kannada and Malayalam. Meta has made Muse Voice Transcribe available through its Meta Model API, allowing developers to use it in their own apps and services. Meta also said that its AI for Max and Muse Code products already use the model for dictation. The API costs 3 dollars which is around 300 Rupees in Indian Currency for 1000 minutes of audio. This works out to about 0.18 dollar which is roughly around 17 Rupees per hour of audio.

Overall, Meta says Muse Voice Transcribe is built to make real time speech transcription easier especially for conversations where people speak different languages. In India, people often switch between English and regional languages while speaking. Because of this, Muse Voice Transcribe could be especially useful in the Indian context. It can make real time transcription easier and more useful for multilingual conversations. The model provides streaming transcription, which means the text appears as the person speaks instead of waiting for the whole recording to finish. Meta says the model has been trained on more than 70 languages, while 25 languages were tested and validated at the time of its launch. The company also claimed in its blog post that Muse Voice Transcribe ranked first on the Artificial Analysis streaming speech to text leaderboard as of September 1^st 2026.

Lastly, the launch comes at a time when voice based AI is becoming more common in areas such as transcription, dictation, coding and voice assistants. Support for code switching and more regional languages could help these systems understand how people actually communicate especially in multilingual countries like India. Meta has also shared more technical details about Muse Voice Transcribe in a research blog published along with the launch.

Source : The Indian Express, and Meta

For more information on IT Services, Web Applications & Support kindly call or WhatsApp at +91-9733733000 or you can visit https://www.technodg.com