New Delhi, November 12, 2026: Chariot, an AI research lab founded by former Google DeepMind and Meta FAIR researchers and one of the first four teams selected under the IndiaAI Mission, has launched Chariot-1-TTS-Multilingual, a new multilingual text-to-speech model supporting 11 Indic languages and English.
The model offers more than 50 voices, code-mixing within sentences, and instant voice cloning from a few seconds of reference audio. It can generate speech from up to 1,000 characters in a single request, aiming to simplify the process of converting text to speech by embedding contextual understanding directly into the model.
Model ranks first in benchmarks
Chariot-1-TTS-Multilingual ranked first on Hindi and Indic accuracy benchmarks in the company’s launch comparison, outperforming selected model versions from ElevenLabs, Cartesia, Microsoft, OpenAI, AWS, and Sarvam. The model achieved a character error rate of 7.60% on Hindi and 6.47% across nine other Indic languages in intelligibility tests, and recorded the highest English verbalization accuracy at 93.78% on a benchmark of over 1,000 sentences spanning 40 categories. It also led in plain-English NISQA scores, an automated speech quality metric.
Contextual speech generation features
Chariot-1-TTS-Multilingual is designed to interpret context-sensitive expressions such as payment amounts, grades, blood groups, dates, abbreviations, and technical terms, reducing the need for custom normalization rules or engineered prompts. The model’s evaluation framework tests intelligibility, reasoning accuracy, and quality, including code-mixed sentences and technical expressions.
Applications for businesses and developers
The model is targeted at businesses and developers building voice agents, customer-service systems, learning applications, and multilingual content. Organisations can select from the voice catalogue or create new voices from short reference recordings for use cases such as payment reminders, narration, and public helplines.
Chariot’s research and leadership
Parth Sarthi, Co-Founder of Chariot, said the model’s contextual understanding allows developers to focus on building effective voice experiences rather than writing normalization rules. Suvrat Bhooshan, Co-Founder, added that the company’s ambition is to develop foundation models that understand, generate, and reason in speech, with a focus on supporting the linguistic diversity of India.
Chariot continues to expand its Chariot-1 model family across speech recognition and generation, with a research approach that connects understanding and speaking capabilities in AI.
Disclaimer: This article is based on a press release provided by the concerned organisation and has been edited for style and clarity with the assistance of AI tools. BusinessFortnight does not verify, endorse, or take responsibility for the claims, figures, or statements made in the original release.
