Sarvam AI expands portfolio with 'Sarvam Samvaad'; can deploy AI agents in 11 Indian languages

Homegrown AI startup Sarvam AI has launched Sarvam Samvaad, a new conversational AI platform for businesses to build, test, and deploy AI agents across 11 Indian languages. With Sarvam Samvaad, the company says it aims to transform customer experience by enabling seamless interactions across various channels.

Sarvam says the new platform can help businesses power interactions across telephone, WhatsApp, web, and apps; handle complex phrases, alpha-numerics, and proper nouns with precision; listen in on every conversation and discover deep insight; and scale support with pricing built for India. With this, the company aims to deliver "production-ready agents" to enterprises so they can move from pilot to full-scale deployment in just days.

Sarvam Samvaad's launch follows the release of Sarvam-Translate, an open-weight model that translates text across 22 Indian languages, with support for long-form text and the ability to handle diverse formats, contexts, and styles, which was launched on June 7, 2025.

The company claims Sarvam-Translate can handle the complexities of real-world content. The model can translate web pages while preserving structure, interpreting idioms and slang, can translate social media posts with emojis, and preserve syntax when translating code files, it adds.

Sarvam claims Sarvam-Translate has achieved "significant improvements" over much larger models on key Indian language translation benchmarks. "In human evaluations, conducted by language experts, Sarvam-Translate consistently outperformed Gemma 2-27B-IT, Llama4 Scout, and Llama-3.1-405B. In automated evaluation on the ability to conform to structured long-form content, Sarvam-Translate shows high accuracy (> 4.9 over 5) for 15 languages," the company adds.

Founded in July 2023 by Vivek Raghavan and Pratyush Kumar, Sarvam aims to make Generative AI accessible at scale in India. In April 2025, under the IndiaAI Mission, the Centre selected Sarvam to build India’s sovereign LLM, for which it will receive dedicated compute resources to build an indigenous foundational model from scratch.

Just last month, Sarvam AI launched India's first sovereign LLM, Sarvam-M', while claiming to have an edge over big AI models like LLaMA-3 and Gemma 3 on key benchmarks. The flagship AI model developed by Bengaluru-based Sarvam AI achieved a new benchmark across a range of Indian languages, math, and programming tasks for a model of its size. The LLM is built for versatility and designed to support a wide range of applications, including conversational agents, translation, educational tools, and much more.

The AI model shows 20% average improvement over the base model on Indian language benchmarks, 21.6% enhancement on math-related tasks, and 17.6% improvement in coding benchmarks. The company claims performance gains are even more impressive at the intersection of Indian languages and mathematics, with an outstanding +86% improvement in romanised Indian language GSM-8K benchmarks, highlighting its capability in handling code-mixed language scenarios. The key features of Sarvam-M include a hybrid thinking mode, advanced Indic skills, superior reasoning, and seamless chatting in a multilingual conversation experience.