\u00a0<\/p>\n
India\u2019s growing artificial intelligence<\/strong><\/a> ecosystem received a major boost as Sarvam AI<\/strong><\/a> unveiled two new large language models (LLMs) \u2014 Sarvam-30B and Sarvam-105B<\/strong> \u2014 at the India AI Impact Summit on 18 February. The announcement marks a significant milestone in the country\u2019s push toward sovereign AI infrastructure and homegrown foundation models.<\/span><\/p>\n \u00a0<\/p>\n Positioned as open-source alternatives to foreign AI systems, the two models are designed to deliver high-quality performance while improving efficiency and reducing token usage \u2014 a critical factor in lowering operational costs for enterprises and developers.<\/span><\/p>\n \u00a0<\/p>\n The launch comes at a time when India is accelerating its efforts to build domestic AI capabilities under the government-backed IndiaAI Mission. With growing concerns around data security, infrastructure costs, and reliance on overseas technology providers, initiatives like Sarvam AI\u2019s are seen as strategic steps toward technological self-reliance.<\/span><\/p>\n \u00a0<\/p>\n By releasing both models as open source, Sarvam AI aims to foster adoption among:<\/strong><\/p>\n \u00a0<\/p>\n Developers building AI-powered applications<\/span><\/p>\n Enterprises seeking localized AI solutions<\/span><\/p>\n Government agencies prioritizing data sovereignty<\/span><\/p>\n Research institutions advancing AI innovation<\/span><\/p>\n \u00a0<\/p>\n This approach encourages transparency, collaboration, and customization \u2014 elements that are crucial for scaling AI within regulated and multilingual markets like India.<\/span><\/p>\n \u00a0<\/p>\n According to Pratyush Kumar<\/strong><\/a>, Founder and CEO of Sarvam AI<\/a><\/strong>, the models were developed entirely from the ground up.<\/span><\/p>\n \u00a0<\/p>\n He emphasized that the systems were created without external dataset dependencies, underscoring the company\u2019s focus on developing indigenous AI capabilities. This foundational independence aligns with broader national goals of reducing technological reliance on foreign ecosystems.<\/span><\/p>\n \u00a0<\/p>\n The Sarvam-30B model is the smaller of the two but has been engineered with efficiency at its core.<\/strong><\/p>\n \u00a0<\/p>\n \u00a0<\/p>\n Pre-trained on 16 trillion tokens<\/span><\/p>\n Supports a 32,000-token context window<\/span><\/p>\n Designed to generate high-quality responses using fewer tokens<\/span><\/p>\n \u00a0<\/p>\n Token efficiency is a critical metric in large language models because it directly impacts compute requirements and operational costs. By optimizing response quality while reducing token consumption, Sarvam-30B aims to make large-scale AI deployment more cost-effective for enterprises.<\/span><\/p>\n \u00a0<\/p>\n During the summit, Sarvam AI presented benchmark results indicating that Sarvam-30B outperformed several well-known AI systems in specific evaluation categories.<\/span><\/p>\n \u00a0<\/p>\n The larger Sarvam-105B model significantly expands capabilities with a 128,000-token context window.<\/span><\/p>\n \u00a0<\/p>\n \u00a0<\/p>\n Ability to process longer documents<\/span><\/p>\n Handle extended multi-turn conversations<\/span><\/p>\n Support enterprise-scale knowledge management tasks<\/span><\/p>\n \u00a0<\/p>\n According to the company, Sarvam-105B performs on par with other frontier models of comparable size \u2014 including both open-source and proprietary systems. The expanded context window makes it particularly suitable for sectors such as legal services, research, governance, and enterprise documentation.<\/span><\/p>\n \u00a0<\/p>\n One of the most compelling moments at the launch was a live demonstration showcasing a chatbot named \u201cVikram\u201d running on a basic feature phone equipped with a physical keypad.<\/span><\/p>\n \u00a0<\/p>\n The demonstration highlighted:<\/strong><\/p>\n \u00a0<\/p>\n Multi-language conversational capability<\/span><\/p>\n Operation in low-resource hardware environments<\/span><\/p>\n Practical applicability beyond high-end smartphones<\/span><\/p>\n \u00a0<\/p>\n The chatbot interacted in Hindi and Punjabi, reflecting the company\u2019s focus on supporting India\u2019s linguistic diversity.<\/span><\/p>\n \u00a0<\/p>\n The name \u201cVikram\u201d pays tribute to Vikram Sarabhai, the renowned Indian physicist widely regarded as the father of India\u2019s space programme. The naming underscores the broader theme of technological self-reliance and national innovation.<\/span><\/p>\n \u00a0<\/p>\n By making Sarvam-30B and Sarvam-105B open source, Sarvam AI is positioning itself within a growing global movement toward transparent AI development.<\/span><\/p>\n \u00a0<\/p>\n Open-source models provide:<\/strong><\/p>\n \u00a0<\/p>\n Greater customization for enterprise needs<\/span><\/p>\n Improved auditability and compliance<\/span><\/p>\n Faster innovation through community contributions<\/span><\/p>\n Reduced vendor lock-in<\/span><\/p>\n \u00a0<\/p>\n For India, this could translate into stronger AI ecosystems tailored to local languages, governance requirements, and business environments.<\/span><\/p>\n \u00a0<\/p>\n Sarvam AI joins a growing group of domestic startups working to build large-scale AI models optimized for India\u2019s multilingual and enterprise landscape.<\/span><\/p>\n \u00a0<\/p>\n The launch signals several broader trends:<\/strong><\/p>\n \u00a0<\/p>\n Increased focus on AI sovereignty<\/span><\/p>\n Investment in domestic compute and infrastructure<\/span><\/p>\n Expansion of open-source AI frameworks<\/span><\/p>\n Development of AI systems tailored to emerging markets<\/span><\/p>\n \u00a0<\/p>\n As India aims to position itself as a global AI hub, the introduction of high-parameter, open-source foundation models marks an important step in that direction.<\/span><\/p>\n \u00a0<\/p>\n The unveiling of Sarvam-30B and Sarvam-105B reflects a strategic shift toward building AI systems that are not only powerful but also efficient, accessible, and locally governed.<\/span><\/p>\n \u00a0<\/p>\n If widely adopted, these models could accelerate AI innovation across sectors ranging from education and governance to finance and telecommunications. More importantly, they reinforce India\u2019s ambition to shape its own AI future rather than relying exclusively on global technology giants.<\/span><\/p>\n \u00a0<\/p>\n With open-source accessibility, extended context capabilities, and a focus on token efficiency, Sarvam AI\u2019s latest models represent a significant contribution to India\u2019s evolving artificial intelligence landscape.<\/span><\/p>\n \u00a0<\/p>\n Latest Technology News | Updates<\/a> | Latest Electric Vehicle News | Updates<\/a> | Electronics News<\/a> | Mobile News | Updates<\/a> | Software Updates<\/a><\/p>\n \n Facebook<\/a> | Twitter<\/a> | WhatsApp Channel<\/a> | Instagram<\/a> | Telegram<\/a> | Threads<\/a> | LinkedIn<\/a> | YouTube<\/a><\/p>\n \u00a0<\/p>\n Stay informed, Stay Connected!<\/p>\n The post Sarvam AI Launches 30B and 105B Open-Source LLMs at India AI Impact Summit, Marking Major Push for Homegrown AI Development<\/a> appeared first on Before You Take<\/a>.<\/p>","protected":false},"excerpt":{"rendered":" \u00a0 India\u2019s growing artificial intelligence ecosystem received a major boost as Sarvam AI unveiled two new large language models (LLMs) \u2014 Sarvam-30B and Sarvam-105B \u2014 at the India AI Impact Summit on 18 February. The announcement marks a significant milestone in the country\u2019s Continue reading Sarvam AI Launches 30B and 105B Open-Source LLMs at India AI Impact Summit, Marking Major Push for Homegrown AI Development<\/span>Strengthening India\u2019s Sovereign AI Ambitions<\/h2>\n
Built from Scratch: A Fully Indigenous Approach<\/h2>\n
Sarvam-30B: Efficiency-Focused Large Language Model<\/h2>\n
Key Highlights:<\/h3>\n
Sarvam-105B: Advanced Model for Long-Context Intelligence<\/h2>\n
What This Means:<\/h3>\n
Real-World Demonstration: AI on a Feature Phone<\/h2>\n
Open Source Strategy: Driving Adoption and Innovation<\/h2>\n
Implications for India\u2019s AI Ecosystem<\/h2>\n
The Road Ahead<\/h1>\n
Follow Before You Take<\/a> on:\u00a0<\/span><\/h3>\n