Seetha Mahalaxmi Healthcare (SML) and 3AI Holding jointly unveiled Hanooman AI on Friday. On February 21, the AI was allegedly demonstrated at an event in Mumbai, leading to the announcement of the large language model (LLM) in India.
It is said that the AI model was created with the Indian population in mind. It can translate text and is compatible with 98 global languages, including 12 Indian languages. All users can presently access the Hanooman AI platform for free, but in the future, the company intends to develop a premium version that requires a paid subscription.
Features of Hanooman AI
The developers claim that the chatbot is intended to work in a number of significant industries, such as healthcare, government, financial services, and education.
Hanooman AI can answer to users from a wide variety in their own language thanks to support for 12 Indian languages and a total of 98 worldwide languages. The AI is multilingual and can speak a variety of languages, including English, French, Spanish, Italian, German, Japanese, Russian, and Chinese.
The chatbot essentially does everything that OpenAI’s ChatGPT or Google’s Geminican do. It can produce text, converse, offer recipes, and respond to questions. The amount of the training data set and the AI model’s parameters are not disclosed by the company. Hanooman, however, “integrates specialized LLMs with an integration synthesis matrix, facilitating the delivery of clear, adaptive insights and seamless complex data transformation into actionable intelligence,” according to the company’s explanation.
After testing the chatbot, it was discovered that it performs passably on simple text-based generative AI tasks. Most notably, it lacks internet access and is not multimodal. This indicates that it lacks real-time information and is unable to create or accept photos as input. Hanooman AI informed us that its knowledge cut-off date was April 10, 2022 when we asked.
It supports multilingual communication between users. We tested Bengali, Hindi, and English. It responded succinctly in Bengali, but occasionally had trouble with the former. In certain instances, the replies lacked coherence and clarity, or they were unable to comprehend the input provided.
