India's Sarvam AI Raising $300. 350 Million at $1.5 Billion Valuation to Challenge US and Chinese AI Leaders

Sarvam AI, India's most prominent homegrown artificial intelligence startup, is close to raising between $300 million and $350 million in a funding round that would value the company at $1.5 billion, Bloomberg reported, as the Bengaluru-based company positions itself as a domestic AI champion capable of competing with leading models from the United States and China. Sarvam has focused on building AI models optimized for Indian languages, dialects, and use cases, arguing that the dominant English-centric models from OpenAI, Anthropic, and Google are poorly suited to the linguistic and cultural diversity of a country with 22 officially recognized languages and hundreds of additional dialects spoken across a population of 1.4 billion. The company's technical approach centers on building foundation models and fine-tuned variants specifically for Indic languages, as well as AI tools for sectors with high India-specific demand including healthcare, agriculture, financial services, and government services. Sarvam has worked with Indian government agencies on AI deployments and has positioned itself as a trusted domestic provider for sensitive public sector applications. The $1.5 billion valuation would make Sarvam one of the most valuable AI startups in Asia outside of China, reflecting both investor confidence in the India AI market and growing government support for domestic technology champions. India has announced significant AI investment initiatives and has expressed preference for domestic providers in public sector AI deployments. The funding round investor composition was not disclosed. Previous Sarvam backers have included Lightspeed India and other venture funds with India focus. The raise comes as global competition to build language-specific AI models has intensified, with European, Japanese, Korean, and Arabic-language AI projects all gaining momentum as local governments and enterprises seek alternatives to US-dominated AI infrastructure.