First open platform to benchmark AI image generators through head-to-head human voting with tamper-proof audit trail ...

B and 105B parameters—under an Apache 2.0 license, aiming to support India’s sovereign AI push and compete with global LLMs.

Sridhar Vembu, founder of Zoho, has backed Sarvam AI’s strategy following the launch of its Sarvam 30B and 105B models, ...

Whether it is a 0.8B model running on a smartphone or a 9B model powering a coding terminal, the Qwen3.5 series is effectively democratizing the "agentic era." ...

Let’s take a look at how data infrastructure stocks fared in Q4, starting with Oracle (NYSE:ORCL). Generating insights from system level data is an increasing priority for most businesses, but to do ...

Looking for a genuinely engaging motorcycle under Rs 1.50 lakh in India now means balancing performance, weight and ...

UniPat AI 开源了他们的一个科研 agent 模型 UniScientist，后训练自 Qwen3，30B参数，3B激活在 FrontierScience-Research 榜单上拿了 28.3 分，多跑几次聚合后 33.3，略微还超过了 GPT-5.4（33.0分）https://github.com/UniPat-AI/UniScientist但....别急着下结论，虽然是老朋友，我也 ...

The 30-billion- and 105-billion-parameter models are available for download under an open-source licence via AIKosh and ...

This leap is made possible by near-lossless accuracy under 4-bit weight and KV cache quantization, allowing developers to process massive datasets without server-grade infrastructure.

Sarvam AI's 105B is a genuine engineering achievement. But India still lacks a trusted, independent institution to verify whether its sovereign models perform as claimed.

Check out the FASTER Awards 2026 Winners, starting from TVS Apache RTX 300, Mahindra XEV 9e, and Honda Activa.

Sarvam AI has released two open-source AI models trained in India. Zoho’s Sridhar Vembu said the development highlights the ...