As artificial intelligence (AI) becomes the foundation of the new digital economy, governments around the world are investing in building their own AI infrastructure. While global technology leaders such as OpenAI, Google, Anthropic, and Meta continue advancing frontier AI models, India is pursuing a different objective: building sovereign AI that understands the country’s languages, culture, regulations, and enterprise needs.
At the center of this effort is Sarvam AI, a Bengaluru-based startup that has rapidly emerged as one of India’s leading AI companies. Founded in 2023, the company focuses on developing foundation models, multilingual speech technologies, and enterprise AI solutions for India. In 2026, Sarvam AI joined India’s unicorn club after raising US $234 million in the first close of its Series B funding round at a valuation of US$1.5 billion.
Unlike many startups that build application-layer products on top of open-source AI models, Sarvam AI is developing the underlying AI infrastructure itself. The company is building large language models (LLMs), speech AI, translation systems, and AI agents to support India’s sovereign AI ecosystem. As governments and enterprises increasingly seek secure, homegrown, and multilingual AI solutions, Sarvam AI is positioning itself as a key player in India’s AI transformation.
Sarvam AI – Snapshot (2026)
| Particular | Details |
| Company | Sarvam AI |
| Founded | 2023 |
| Headquarters | Bengaluru, India |
| Founders | Dr. Vivek Raghavan and Dr. Pratyush Kumar |
| Industry | Generative AI, Foundation Models |
| Latest Valuation | $1.5 Billion |
| Latest Funding | $234 Million (Series B First Close) |
| Status | AI Unicorn |
| Focus | Sovereign AI, Enterprise AI, Foundation Models |
| Core Products | Sarvam 30B, Sarvam 105B, Saaras V3, Bulbul V3, Sarvam Vision, Indus AI |
Meet the Founders Behind Sarvam AI
Sarvam AI was founded by two leaders with deep expertise in artificial intelligence and India’s digital public infrastructure.
Dr. Vivek Raghavan has played an important role in advancing India’s digital public infrastructure initiatives. Before co-founding Sarvam AI, he was involved in projects such as AI4Bharat, which focuses on developing AI technologies for Indian languages. His experience in building large-scale digital platforms has helped shape Sarvam AI’s approach to addressing India’s unique technological and linguistic requirements.
Co-founder Dr. Pratyush Kumar is an AI researcher specializing in machine learning, multilingual language models, and speech technologies. His research guides Sarvam AI’s technical strategy, particularly in developing foundation models tailored for Indian languages and enterprise use cases.
The founders are united in the conviction that India should not be reliant on imported AI models but must build its own groundbreaking AI infrastructure.
Why Was Sarvam AI Founded?
The global explosion of generative AI is changing the way businesses operate. But almost all of the dominant AI models are built outside India and predominantly trained on English-language data.
This creates several challenges:
- Limited AI support for India’s multilingual population.
- Reliance on overseas AI systems.
- Concerns over data sovereignty.
- Regulatory implications and security issues.
- Increased expense for companies pursuing regional AI solutions.
To address these challenges, Sarvam AI is developing AI systems specifically designed for India. Rather than adapting models developed in other countries, the company is building multilingual foundation models that are trained to understand Indian languages, culture, and enterprise requirements from the ground up.
What Is Sovereign AI?
One of Sarvam AI’s defining strengths is its focus on sovereign AI.
Sovereign AI refers to artificial intelligence that is developed, trained, deployed, and governed within a country’s own technological ecosystem. This approach reduces dependence on foreign AI providers while giving governments and organizations greater control over critical technologies, data, infrastructure, and regulatory compliance.
For India, this means developing AI models that:
- Choose from various Indian Languages.
- Stay powered by an in-home AI system.
- Comply with the Indian regulations that are in place.
- Enhance data privacy and security.
- Limit the long-term reliance on foreign AI vendors.
This vision closely aligns with the Government of India’s IndiaAI Mission, which aims to accelerate AI innovation and strengthen indigenous AI capabilities across the country.
Funding Journey: From Startup to Unicorn
Sarvam AI’s funding journey reflects the growing confidence investors have in India’s AI ecosystem.
In 2023, the company raised approximately US$41 million in its Series A funding round from investors including Lightspeed, Peak XV Partners, and Khosla Ventures. The funding enabled Sarvam AI to expand its engineering team, develop its core language models, and accelerate product development.
Sarvam AI reached a major milestone in the first half of 2026 when it announced the first close of its Series B funding round, raising US$234 million at a valuation of US$1.5 billion. Strategically led by HCLTech, the round is expected to reach approximately US$300 million in total funding. The funding officially established Sarvam AI as a unicorn and marked one of India’s largest investments in a foundational AI company.
Beyond its size, the funding reflects growing investor confidence in India’s ability to develop world-class, indigenous AI technologies and reduce dependence on foreign AI platforms.
Powering India’s AI Ambitions Through the IndiaAI Mission
Sarvam AI’s growth is closely aligned with the Government of India’s efforts to build indigenous AI capabilities.
Under the IndiaAI Mission, the company was selected to help develop India’s sovereign foundation models. As part of the initiative, Sarvam AI has been allocated 4,096 NVIDIA H100 GPUs along with nearly ₹99 crore in GPU subsidies, making it one of the largest beneficiaries of the government’s AI infrastructure program. The IndiaAI Mission, backed by a budget of more than ₹10,000 crore, reflects India’s commitment to building a world-class AI ecosystem.
Rather than supporting a single startup, the initiative aims to establish foundational AI infrastructure that can serve government agencies, enterprises, developers, and hundreds of millions of users across India’s diverse linguistic landscape.
Backed by government support, advanced computing infrastructure, and strong in-house AI research, Sarvam AI is well positioned to become one of the leading contributors to India’s AI ecosystem.
Sarvam AI’s Business Model
Unlike many consumer AI companies that primarily rely on chatbot subscription revenue, Sarvam AI follows an enterprise-first business model.
The company develops core AI models and commercializes them through APIs, enterprise software, AI infrastructure services, and government partnerships.
Its revenue streams include:
- Application Programming Interfaces (APIs) for Foundation Models: Businesses access Sarvam AI’s large language models through APIs to build AI-powered applications and services.
- Enterprise AI Platforms: Companies deploy customized AI assistants, multilingual processing solutions, and workflow automation tools tailored to their business requirements.
- Speech and Voice AI: Sarvam provides speech-to-text, text-to-speech, and conversational AI APIs for customer support, finance, health care, and telecom industries.
- Translation and Indic Language AI: Businesses leverage Sarvam’s translation support to create products and services that cater to the multilingual nature of India.
- Government and Public Sector Projects: Participating in the IndiaAI Mission, Sarvam supports developing India’s AI infrastructure and sovereign AI.
- Developer Ecosystem: The company provides developers and startups with APIs, technical support, documentation, and startup programs to encourage widespread adoption of its AI models.
This kind of diversified approach will allow Sarvam AI to cater to both public and private industries, as well as multiple revenue streams in the future.
Products Powering India’s AI Ecosystem
Sarvam AI has developed a broad portfolio of AI products designed to serve developers, enterprises, and government organizations across multiple use cases.
Sarvam 105B
Sarvam 105B is the company’s premier large language model designed for advanced reasoning, multilingual content, and enterprise-level AI use cases. It features a 128K-token context window, enabling it to handle lengthy documents and intricate business processes. An open-weights version of the model is also available, allowing developers and enterprises to customize it for their specific requirements.
Sarvam 30B
Sarvam 30B is a lightweight multilingual large language model optimized for efficient deployment while delivering strong performance across Indian languages. Its lower computational requirements make it well-suited for enterprises and developers seeking cost-effective AI solutions.
Saaras V3
Saaras V3 is a speech recognition and automatic speech-to-text conversion voice engine for Indian languages. Powered by this engine, businesses can develop voice-first applications for customer care, healthcare, and digital services.
Bulbul V3
Bulbul V3 – Text-to-speech (TTS) with a natural voice as an offering supporting many Indian languages, for organizations to develop localized voice assistants and conversational AI experiences, among others.
Sarvam Vision
Sarvam Vision integrates optical character recognition (OCR) with multimodal AI to read out, comprehend, and act on data from documents, images, and scanned files, making it useful for bank, insurance, and government processes.
Indus AI
Indus AI is the consumer-facing AI assistant for the company and provides Sarvam with the first real test of multilingual conversational chat, knowledge search, and an AI assistant in day-to-day life.
Together, these products position Sarvam AI as a full-stack AI company serving developers, enterprises, and public-sector organizations.
Open-Source Strategy: Building an AI Ecosystem
Whereas many other leading AI companies around the world keep their state-of-the-art models closed, Sarvam has unveiled open weights to facilitate research as well as corporate integration and development. This decision allows organizations to operate AI directly on their premises without being reliant on proprietary commercial offerings. (timesofindia.indiatimes.com)
In India, open-source AI is also part of the larger goal of creating a self-sustaining AI ecosystem that may develop through partnerships across startups, academia, large businesses and government organizations.
Competitive Landscape
Sarvam AI is operating in a highly competitive technology market worldwide.
Leading the way on frontier AI, groups from OpenAI and Google to Anthropic, Meta and DeepSeek break new ground. Rather than go head-to-head in a broad range of categories, Sarvam concentrates on where it has a clear edge.
Its strengths include:
- Hard to optimize for Indian languages.
- AI models targeted towards local businesses.
- Voice AI and speech technologies.
- Risk of an ever-expanding government-funded sovereign AI program
- Excellent knowledge and comprehension of the regulatory frameworks and multi-lingual environment in India.
This focused positioning enables Sarvam AI to compete by addressing India’s unique AI requirements rather than directly replicating the strategies of global AI companies.
Challenges Ahead
However, for all its rapid expansion, Sarvam AI still has a few hurdles to overcome:
- Training large language models (LLMs) is highly capital-intensive and requires substantial computing resources, including continuous investment in GPUs, cloud infrastructure, and model maintenance.
- The company also faces intense competition from global AI firms, requiring continuous innovation, talent acquisition, and technological advancement to remain competitive.
- Commercializing foundation models remains another significant challenge.
- Although demand for AI continues to grow, AI startups must convert technological leadership into sustainable and recurring revenue streams.
- Maintaining leadership in multilingual AI while keeping pace with rapidly evolving frontier models will require sustained research and long-term investment.
The Future of Sarvam AI
Sarvam AI is entering its next growth phase.
The company plans to accelerate enterprise AI adoption, develop multilingual foundation models, enhance voice and multimodal AI capabilities, and strengthen India’s AI ecosystem through the IndiaAI Mission.
As demand for sovereign AI increases across governments and regulated industries, Sarvam AI is well positioned to play an even greater role in India’s AI ecosystem.
Rather than building another chatbot, the company aims to develop the foundational AI technologies that enable businesses, developers, and public institutions to create AI solutions tailored to India’s unique requirements.
Conclusion
Sarvam AI represents the broader transformation taking place within India’s AI ecosystem. Rather than relying solely on AI technologies developed overseas, the company is helping build indigenous foundation models, multilingual AI systems, and sovereign AI infrastructure designed specifically for India’s needs.
From achieving unicorn status to becoming a key participant in the IndiaAI Mission, Sarvam AI has rapidly established itself as one of India’s leading AI startups.
Its combination of foundation models, speech AI, enterprise solutions, and government collaborations demonstrates that India’s AI ambitions extend well beyond building consumer applications.
As artificial intelligence becomes an increasingly important driver of the global digital economy, Sarvam AI is well positioned to contribute to the development of India’s next generation of AI infrastructure.
1. What is Sarvam AI used for?
Sarvam AI is used for developing AI applications across multiple Indian languages, enterprise chatbots, voice assistants, translation applications, speech recognition applications, document processing applications and AI agents. Its foundation models enable businesses and government agencies to implement AI applications around Indian languages and use cases.
2. How is Sarvam AI different from ChatGPT?
Unlike ChatGPT, an open-domain conversational AI by OpenAI for a global audience, Sarvam AI builds India-first foundation models tuned for Indian languages, enterprise applications, government AI initiatives, and sovereign AI. The company also focuses on speech AI, translation AI, and enterprise APIs rather than a consumer-focused conversational AI platform.
3. Is Sarvam AI India’s answer to OpenAI?
Sarvam AI is commonly recognized as being one of India’s premier sovereign AI startups, as it is building native foundation models and an AI infrastructure. Rather than directly competing with OpenAI across every category, the company focuses on solving India-specific challenges such as multilingual AI, enterprise adoption, and government-led AI initiatives.
4. How does Sarvam AI make money?
Sarvam AI generates revenue through foundation model APIs, enterprise AI platforms, speech and voice AI services, multilingual translation solutions, custom AI projects, and government partnerships. Its enterprise-focused business model serves organizations, developers, startups, and public-sector institutions.
5. What makes Sarvam AI stand out as one of the leading AI startups in India?
Sarvam AI stands out because of its rapid growth, strong investor backing, focus on sovereign AI, and active participation in the IndiaAI Mission. Its emphasis on developing foundation models for Indian languages, enterprise AI solutions, and indigenous AI infrastructure has positioned it as one of India’s leading AI startups.
