Sarvam AI

Sarvam AI is an Indian artificial intelligence company founded in August 2023 by Dr. Vivek Raghavan and Dr. Pratyush Kumar, both IIT alumni. Headquartered in Bengaluru, the company builds large language models, speech systems, and enterprise AI products specifically designed for India's linguistic diversity.

Models

In February 2026, Sarvam open-sourced two flagship models at the India AI Impact Summit: Sarvam 30B (30 billion parameters) and Sarvam 105B (105 billion parameters, using a Mixture-of-Experts architecture with 10.3 billion active parameters). Both were trained from scratch on Indian language datasets and support reasoning, multilingual tasks, mathematics, and coding. The company also released Sarvam-1, a 2-billion parameter model optimized for 10 Indic languages, along with dedicated text-to-speech, speech-to-text, and vision models.

Products

Sarvam's product lineup includes Indus, a consumer AI chatbot available in 22 Indian languages; Sarvam Translate, an open-weights translation model covering all 22 officially recognized Indian languages; Sarvam Agents, voice-enabled multilingual business agents deployable via phone, WhatsApp, or in-app; and Sarvam Vision, a vision-language model for document understanding and OCR in Indian scripts.

The company also announced Sarvam Kaze, an AI-powered wearable glass supporting real-time voice interaction and translation in 10+ Indian languages, planned for launch in May 2026.

Funding and Government Support

Sarvam raised approximately $41 million in a combined seed and Series A round in December 2023, led by Lightspeed Venture Partners with participation from Peak XV Partners and Khosla Ventures. Total funding stands at around $54 million. In April 2025, the Ministry of Electronics and Information Technology (MeitY) selected Sarvam AI under the IndiaAI Mission to develop indigenous foundational models, providing access to government-supported GPU infrastructure.

Sarvam positions itself as a sovereign AI alternative — building models trained on Indian data, optimized for Indian languages, and deployable on Indian infrastructure.