Sarvam AI is reportedly in talks to raise $250–300 million at a potential $3 billion valuation.
Indian artificial intelligence startup Sarvam AI is reportedly in talks to raise $250 million to $300 million in fresh funding at a potential valuation of $3 billion, according to a report published by The Economic Times on October 13, 2026.
If the deal goes through at the proposed valuation, Sarvam could become India’s most valuable AI startup. The valuation would nearly double its previous $1.5 billion valuation in about four months.
The discussions are reportedly at an early stage, with large domestic companies and institutions among the potential participants. The final funding amount and valuation could change as negotiations progress. The latest funding discussions have not been announced as a completed transaction by the company.
Previous Funding Round Valued Sarvam at $1.5 Billion
Sarvam announced in June 2026 that it had raised $234 million in the first close of its targeted $300 million Series B funding round at a post-money valuation of $1.5 billion. The round was subsequently expanded through additional investment, taking the total funding beyond $300 million.
HCLTech committed $150 million as the lead strategic investor, while Bessemer Venture Partners and existing investors Khosla Ventures and Peak XV Partners also participated. Subsequent reporting described a $75 million extension involving investors including Nvidia and IndiGo Ventures.
The new funding discussions come as Sarvam seeks to expand its research capabilities, computing infrastructure and commercial deployments. Training advanced AI models requires substantial computing power, specialised chips and research talent, making access to capital important for companies developing their own foundational models.
What Does Sarvam AI Do?
Founded in 2023 by Pratyush Kumar and Vivek Raghavan, Sarvam AI is a Bengaluru-based company developing artificial intelligence models and applications designed for India’s linguistic and business needs.
Unlike companies that primarily build applications using existing AI models, Sarvam says it develops its own foundational models and works across different parts of the AI stack, including model training, inference infrastructure and enterprise applications.
Its work spans language, speech, vision and document processing. These technologies can be used to develop systems that understand spoken Indian languages, process handwritten records, translate documents and support voice-based interactions.
In its June funding announcement, Sarvam said its Sarvam 105B model matched or outperformed larger reasoning models on certain knowledge, reasoning and agentic benchmarks. It also described Sarvam 30B as a model optimised to run on consumer hardware. These are company-reported performance claims and should be assessed in the context of the specific benchmarks used.
The company is also developing enterprise AI tools for sectors such as banking, insurance, government services and defence. Its approach combines models with applications that organisations can use for specific tasks rather than relying only on general-purpose chatbots.
Focus on Indian Languages and Government Services
India’s linguistic diversity presents both an opportunity and a challenge for AI developers. Many digital services are designed primarily around text-based interfaces, while a large number of people interact more comfortably through spoken languages.
AI systems capable of understanding regional languages and different speech patterns could make digital services more accessible, particularly for users who face difficulties with English-language interfaces.
Sarvam’s speech and voice technologies are intended to address this need. According to its June announcement, the company’s speech models were processing more than half a million hours of audio each month, while its conversational platform handled over two million interactions daily at that time. These figures were reported by the company and have not been independently audited.
Sarvam also develops document-processing technology. The company said its Sarvam Vision system was being used to digitise more than 35 million pages, including insurance forms and legacy land records. Such technology could help organisations process large volumes of documents that are difficult to handle through conventional digital systems.
Its enterprise customers include Life Insurance Corporation of India, SBI Life, IDFC FIRST Bank, Tata Capital and CRED, according to The Economic Times. The company has also announced partnerships involving government departments and technology firms.
Why Sovereign AI Matters for India
The potential funding round also highlights the growing importance of sovereign AI, a term used to describe a country’s ability to develop and operate AI systems using capabilities and infrastructure it can control.
Countries increasingly view AI as strategic technology because it can influence economic productivity, public services, cybersecurity and access to information. Depending heavily on a small number of foreign AI providers can create concerns around costs, data governance, access and technological dependence.
Sarvam’s work is relevant to this discussion because it aims to develop Indian-trained models and infrastructure for domestic organisations. Its partnership with HCLTech and the Odisha government, announced in July, involves plans for an AI data centre in the proposed Odisha Sovereign AI Park, with a planned capital outlay of ₹14,257 crore for the project. This is a planned project investment, not an amount already spent or funding received by Sarvam.
However, developing domestic AI models alone does not guarantee technological independence. Long-term success will also depend on computing capacity, reliable access to advanced chips, research talent, model performance, affordability and adoption by businesses and public institutions.
What Lies Ahead?
The proposed funding could give Sarvam additional resources to compete in a rapidly developing AI market. The company has said its earlier funding would support research into advanced models for agentic applications, coding and cybersecurity, alongside access to computing resources at scale.
Yet, a higher valuation is not the same as higher revenue or profitability. Investors will ultimately assess whether Sarvam can turn its research into products that businesses and governments use consistently, while competing with well-funded international AI developers.
For now, the reported $3 billion valuation remains a potential outcome rather than a completed funding milestone. If the proposed round closes on those terms, it would mark a significant development for India’s homegrown AI industry and its ambitions to build more of its own AI technology.
About the Author
