Indian artificial intelligence startup Sarvam AI is developing AI-powered smart glasses called Sarvam Kaze, designed to support multilingual voice interactions and reduce dependence on cloud-based processing. The wearable device uses the company's Sarvam Edge technology to bring speech recognition, translation and voice generation directly onto the hardware.
The Bengaluru-based company introduced Kaze at the India AI Impact Summit 2026 in New Delhi. The glasses form part of Sarvam's broader efforts to develop AI systems that can operate across Indian languages and on devices with limited connectivity.
Unlike AI applications that depend on remote servers to process user requests, Sarvam Kaze is designed to execute key language functions locally. This approach allows users to interact with AI without continuously relying on smartphones or internet connections.
According to the company, the glasses can recognise speech across 22 Indian languages, enabling hands-free interactions in different linguistic environments. The technology is also designed to translate conversations between Indian languages, allowing users to communicate without manually operating a separate application.
For instance, a conversation in Telugu could be translated into Hindi through the device. Sarvam says the speech recognition and translation processes can operate directly on the glasses without sending audio to cloud servers.
The underlying Sarvam Edge platform combines speech recognition, machine translation and speech synthesis models optimised for local processing. The company says its on-device speech recognition system can transcribe supported languages in real time, while its voice synthesis technology generates spoken responses.
Sarvam claims its speech recognition system can operate with latency below 130 milliseconds, while its voice synthesis technology can begin producing audio in under 60 milliseconds. These performance figures are company-reported and have not been independently verified.
Local processing also changes how user information is handled. When supported AI functions operate entirely on the device, audio and other inputs do not need to be transmitted to external servers for processing.
This architecture could support AI applications in areas with inconsistent connectivity, including educational environments, rural locations and workplaces where reliable internet access is unavailable.
Beyond smart glasses, Sarvam is developing its Edge technology for smartphones, vehicles and other connected devices. The company has outlined applications including voice-driven financial services, multilingual transcription and AI-assisted interactions.
The development comes as technology companies explore wearable devices that integrate AI capabilities into everyday activities. Products such as Meta's Ray-Ban smart glasses have already introduced voice assistants and other AI-enabled features, while Sarvam is focusing on Indian-language support and local processing.
Although Sarvam has demonstrated its underlying technology, detailed commercial specifications, pricing and wider availability of Kaze have not been confirmed. Its performance in everyday use will depend on the capabilities delivered in the final product.
Disclaimer: This article may include information derived from interviews, press releases, public statements, research, company communications and other publicly available or third-party sources. Such material may be summarised, paraphrased or contextualised for journalistic and editorial purposes. All rights in third-party content remain with their respective owners.