India may have arrived late to the AI party, but Indian startups are upping the ante on AI innovation. India ranks third in the number of AI startups followed by the US and China. With only 66 AI startups in India, this number has more than doubled and stands at 240 in 2024.
This ecosystem is thriving thanks to unprecedented innovation. From creating AI models in Indic languages to launching an app store for AI Agents, Indian startups are leading the way, setting the tone for AI implementation. As we look ahead to 2025, here are some Indian AI startups that made it big in 2024.
KOGO AI
Bengaluru-based deep tech startup KOGO AI launched the world’s first AI Agent store in 2024, making AI implementation accessible to businesses of all sizes, across industries. KOGO’s AI AGent Store comes with a unique value proposition — with 100 ready-to-use AI Agents, businesses can deploy AI Agents across their organization in minutes. Businesses can also customize AI Agents without breaking the bank. On the other hand, developers can also leverage KOGO AI OS to build AI Agents, which will be hosted on the AI Agents Store. It’s like an app store, but for AI Agents. KOGO’s AI Agents converse in multiple Indic languages including Urdu, Hindi, and English. The company plans to launch 1 million ready-to-use AI Agents in its store by the end of 2025. By democratizing access to AI, KOGO is leading the charge in AI innovation.
Sarvam AI
Sarvam AI has developed India’s first home-grown, multilingual model — Sarvam 1. Built entirely on NVIDIA technology, Sarvam 1 is a 2-billion-parameter model, trained on 4 trillion tokens. This means that it is four times more efficient than leading English-trained models on Indian language text. Sarvam 1 supports 11 languages, including Bengali, Gujarati, Hindi, Marathi, Malayalam, Kannada, Oriya, Tamil, Telugu, Punjabi, and English. Sarvam 1 is already powering generative AI agents and other applications. The company has also allowed developers to use the base model to build their own AI applications for Indian languages. It is the first AI model trained from scratch in India, making it the only AI startup to have successfully launched LLMs in Indian languages.
Krutrim AI
Competing with Sarvam AI, Krutirm, launched by Ola’s founder Bhavish Aggarwal, specializes in developing large language models in Indian languages. The company raised $50 million earlier this year, and is valued at $1 billion. With plans to accelerate the development of conversational AI in multiple Indian languages, Krutrim AI is at the forefront of driving inclusive AI adoption to ensure it is accessible to everyone in the country.
Qure AI
With over $60 million in funding, Qure AI is an Indian startup that has gained recognition around the world as a top-notch healthcare AI company. Specializing in AI solutions for lung, heart, neuro, and musculoskeletal (MSK) conditions, Qure.ai works towards improving diagnoses and patient care. With the goal of making healthcare accessible to all, Qure.ai leverages deep learning technology to provide automated interpretation of radiology exams like X-rays, CTs, and ultrasound scans, enabling faster diagnosis and speed to treatment.
Kissan AI
From developing India’s first farmer’s AI copilot to training the first vertical-specific language model, KissanAI is at the front and center of driving AI-led innovation in agriculture. The company recently launched Dhenu2, its latest series of agricultural language models, trained on Llama3.1 and Llama3.2 models. Dhenu2 is optimized for India’s unique agricultural practices and to understand the nuances of Indian farmers. This is powering Kissan AI’s copilot, which assists farmers by providing real-time agricultural insights and solutions. It offers personalized recommendations for crop management, pest control, weather forecasting, and soil health. With the goal of making agriculture sustainable and productive, Kissan AI is delivering actionable data and expert advice directly to farmers, helping them make informed decisions and optimize their agricultural practices.
