TNB Tech Minute: Nvidia Releases New Open Source AI Model

Nvidia released a new open-source AI model called Nemotron 3.5 Lightning on Tuesday, which the company says is smaller, cheaper to run, and designed to help enterprises save on token costs while keeping proprietary data out of the hands of AI companies that might one day compete with them. The announcement came a day after Meta released its own open model, Muse Glimmer, which can run on a consumer-grade Mac or PC with a single GPU. Meanwhile, Anthropic is meeting with potential investors to shore up confidence ahead of its expected IPO, and Intel nearly doubled its planned stock offering to $20 billion.

Nvidia's Enterprise-Focused Open-Source Push

Nvidia's Nemotron 3.5 Lightning is an open-source AI model positioned as a more practical alternative to large, expensive systems. The company says it will help enterprises reduce token costs — the per-query fees charged by many AI providers — and protect proprietary data. The model is designed to be run locally or in a controlled environment, which Nvidia argues prevents companies in specialized fields like cybersecurity and material science from sending their intellectual property to AI companies that could eventually become competitors.