According to the developers, the chatbot has been designed to function in several important sectors, including health services, governance, financial services, and education.
Hanooman AI has been launched by 3AI Holding and Sita Mahalakshmi Healthcare (SML). The Indian Large Language Model (LLM) was introduced on February 21 during an event in Mumbai. The AI model has been tailored for Indian users, supporting 12 Indian languages and a total of 98 global languages, along with text translation capabilities. The Hanooman AI platform is currently available for free to all users, with plans for a premium version with a paid subscription in the future. The official handle of the Hanuman AI platform announced the launch on Instagram, stating, “Say hello to Hanooman, India’s first generative AI, eliminating language barriers! Experience seamless communication in your favorite language! Be it Hindi, Tamil, Bengali, or any other language, Hanuman has got it all covered!” Notably, the company has made its chatbot available through both a web client and an Android app.
Features of Hanooman AI
According to the developers, the chatbot has been designed to function in several important sectors, including health services, governance, financial services, and education. With support for 12 Indian languages and a total of 98 global languages, Hanuman AI can respond to users in their preferred language. The AI supports many global languages, including English, French, Spanish, Italian, German, Japanese, Russian, and Chinese.
The chatbot performs similar tasks to OpenAI’s ChatGPT or Google’s Gemini. It can respond to calls, compose texts, share recipes, and engage in conversations. The company has not disclosed information about the size of the training data or the standards used in the AI model. However, it notes that “Hanuman integrates specialized LLM with an integration matrix that facilitates seamless data transformation into clear, adaptive insights and actionable intelligence.”
We tested the chatbot at Gadgets 360 and found that it performs well in basic text-based generative AI tasks. It is not multimodal and does not have internet access, meaning it cannot take or generate photos and lacks real-time information. Users can communicate with it in different languages, including English, Hindi, and Bengali. While its responses were fluent in English and Hindi, it occasionally struggled with Bengali. We observed some instances where the response was unclear or did not address the query effectively. However, like all generative AI, it is expected to improve as it learns.