BharatGen Speech Models
BharatGen’s speech models cover automatic speech recognition and text-to-speech across major Indian languages, built for accuracy, scale and linguistic diversity.
Shrutam2 is an LLM-powered ASR model that transcribes speech, including support for code-mixed speech like Hindi-English. Supported languages: Hindi, Marathi, Tamil, Telugu, Malayalam, Kannada, Odia, Bengali, Urdu, Assamese, Gujarati and Punjabi.
Sooktam2 is a TTS model that synthesises natural, expressive speech using reference-guided voice conditioning, preserving the speaker’s voice. Supported languages: Hindi, Marathi, Gujarati, Tamil, Telugu, Kannada, Bengali, Malayalam, Odia, Urdu, Punjabi and Indian English.

