2 Sources
[1]
Sarvam AI Launches Bulbul, New Voice Model for Indian Market
The model promises lightning-fast performance and custom voices for brands, enhancing user experience with custom solutions. Sarvam AI has unveiled Bulbul, a new speech AI model that supports 11 Indian languages with natural, region-specific accents. It offers authentic, human-like voice interactions. "Natural, familiar speech in 11 Indian languages, with authentic accents that sound just like India," said Sarvam AI in their announcement. Bulbul's focus is to provide voice interactions that are "Not robotic. Not rehearsed. Just real." The system promises lightning-fast performance and custom voices for brands, enhancing user experience with custom solutions. The first version of Bulbul, called Bulbul v1, was released last year in August and featured six distinct voices, each designed to cater to a wide range of communication needs across various industries and contexts. Sarvam AI was recently selected by the Government of India under the IndiaAI Mission to develop India's sovereign Large Language Model (LLM) as part of the effort to create indigenous AI capabilities. The company recently demonstrated its capabilities by translating high-profile content, including the Lex Fridman podcast featuring Prime Minister Narendra Modi, into nine Indian languages Announcing the development, Sarvam AI said, "Building India's sovereign model from the ground up is a crucial step toward Atmanirbhar Bharat." The model will be designed to be fluent in Indian languages, optimised for voice, capable of reasoning, and ready for secure, population-scale deployment. Sarvam AI has proposed the development of a 70-billion parameter multimodal AI model that supports both Indian languages and English, and work on it has already begun. The start-up, selected from 67 applicants, will receive government support in the form of compute resources to build the model from scratch. Founded in 2023 by Dr. Vivek Raghavan and Dr. Pratyush Kumar, both veterans in AI research, public infrastructure, and large-scale systems, the Bengaluru-based company is backed by Lightspeed, Peak XV Partners, and Khosla Ventures.
[2]
Sarvam Unveils New Speech AI Model With 11 Indian Languages Support
The speech model Bubul V2 will offer custom voice options for brands GenAI startup Sarvam AI, recently selected by the Centre to build India's first homegrown LLM, has unveiled a new speech AI model that supports 11 Indian languages, including Punjabi, Marathi, Odiya, Tamil and Bangla, among others. The Bengaluru-based startup took to X to announce Bulbul V2. 'Meet the all-new Bulbul...Natural, familiar speech in 11 Indian languages, with authentic accents that sound just like India," the X post read. Sarvam AI claims that the model has fast performance and a voice that doesn't appear robotic but real. It further says that the model will also offer a custom voice option for brands to represent their distinctive identity. "From lower latency and India-first pricing to wider language, Bulbul sets a new benchmark for Speech AI in India," it said in the post. Notably, Blubul V1 had six voices to suit different needs. The central government picked Sarvam AI to build India's first homegrown sovereign large language model under the IndiaAI Mission. Sarvam last month said in a statement that it will receive a dedicated compute infrastructure to build the model from scratch. The LLM will focus on reasoning, support voice-based tasks and work fluently across Indian languages. It is being designed for secure, population-scale use. The model will be fully developed, deployed and optimised within the country, using local talent and infrastructure.
Share
Copy Link
Sarvam AI launches Bulbul V2, an advanced speech AI model supporting 11 Indian languages with natural accents, promising enhanced user experience and custom voice options for brands.
Sarvam AI, a Bengaluru-based GenAI startup, has unveiled Bulbul V2, an advanced speech AI model that supports 11 Indian languages, marking a significant advancement in voice technology for the Indian market 12. This new model builds upon its predecessor, Bulbul V1, which was released in August last year with six distinct voices 1.
Bulbul V2 now supports a wide array of Indian languages, including Punjabi, Marathi, Odiya, Tamil, and Bangla, among others 2. The model is designed to provide natural, familiar speech with authentic accents that "sound just like India," aiming to offer voice interactions that are "Not robotic. Not rehearsed. Just real" 1.
The new model boasts several improvements over its predecessor:
Sarvam AI's Bulbul V2 is positioned to cater to a wide range of communication needs across various industries and contexts. The model's ability to provide authentic, human-like voice interactions in multiple Indian languages could significantly enhance user experiences in sectors such as customer service, entertainment, and education 1.
The launch of Bulbul V2 comes at a crucial time for Sarvam AI, as the company was recently selected by the Government of India under the IndiaAI Mission to develop India's sovereign Large Language Model (LLM) 12. This initiative is part of the country's efforts to create indigenous AI capabilities and move towards "Atmanirbhar Bharat" (self-reliant India) 1.
Sarvam AI has proposed the development of a 70-billion parameter multimodal AI model that supports both Indian languages and English. The company will receive government support in the form of compute resources to build this model from scratch 1. The LLM is being designed to be:
Founded in 2023 by Dr. Vivek Raghavan and Dr. Pratyush Kumar, both veterans in AI research, public infrastructure, and large-scale systems, Sarvam AI is backed by prominent investors including Lightspeed, Peak XV Partners, and Khosla Ventures 1. The company's selection for the IndiaAI Mission, chosen from among 67 applicants, underscores its potential to contribute significantly to India's AI landscape 12.
Summarized by
Navi
[1]
NASA and IBM have developed Surya, an open-source AI model that can predict solar flares and space weather, potentially improving the protection of Earth's critical infrastructure from solar storms.
5 Sources
Technology
2 hrs ago
5 Sources
Technology
2 hrs ago
Meta introduces an AI-driven voice translation feature for Facebook and Instagram creators, enabling automatic dubbing of content from English to Spanish and vice versa, with plans for future language expansions.
8 Sources
Technology
19 hrs ago
8 Sources
Technology
19 hrs ago
OpenAI CEO Sam Altman reveals plans for GPT-6, focusing on memory capabilities to create more personalized and adaptive AI interactions. The upcoming model aims to remember user preferences and conversations, potentially transforming the relationship between humans and AI.
2 Sources
Technology
19 hrs ago
2 Sources
Technology
19 hrs ago
Chinese AI companies DeepSeek and Baidu are making waves in the global AI landscape with their open-source models, challenging the dominance of Western tech giants and potentially reshaping the AI industry.
2 Sources
Technology
3 hrs ago
2 Sources
Technology
3 hrs ago
A comprehensive look at the emerging phenomenon of 'AI psychosis', its impact on mental health, and the growing concerns among experts and tech leaders about the psychological risks associated with AI chatbots.
3 Sources
Technology
3 hrs ago
3 Sources
Technology
3 hrs ago