Nvidia has been developing its own AI models since 2023; unlike consumer chatbots, these models are designed to help enterprises build AI applications that run on Nvidia chips.
- On Wednesday, Nvidia released Nemotron 3 Super, its most capable open-weight AI model to date.
- Nvidia has earmarked to spend $26 billion on AI models over the next four years.
- Nvidia investors and enthusiasts are prepping for the company’s developers’ conference, which starts next week and features a host of product announcements.
Nvidia Corp. has indeed earmarked $26 billion for developing its own artificial intelligence (AI) models over the next five years, Wired reported on Wednesday. The move could put the chip giant in more direct competition with enterprise AI offerings from OpenAI, Google, and others.
According to Wired, Nvidia is doubling down on open model development, with executives confirming progress previously hinted at in a 2025 regulatory filing. “Nvidia is taking open model development much more seriously. And we are making a lot of progress. It’s in our interest to help the ecosystem develop,” said Bryan Catanzaro, VP of applied deep learning research, according to the report.
The AI chip giant began developing its own large language models (LLMs) – the Nemotron series – in 2023, as part of its push to build a full enterprise AI stack around its GPUs. The models, released with open weights or permissive licenses for research and enterprise use, are designed to help enterprises build, customize, and deploy AI applications on Nvidia hardware and cloud platforms.
Unlike consumer chatbots, Nvidia’s models are aimed at businesses, developers, and cloud providers, enabling companies to build applications for tasks such as customer service, coding, and data analysis, and to run them on Nvidia’s AI infrastructure.
On Wednesday, Nvidia also released Nemotron 3 Super, its most capable open-weight AI model to date. The model size makes it roughly equivalent to the largest version of OpenAI’s GPT-OSS. Unlike the previous Nemotron models, which mainly served as training frameworks or research models, Nemotron-3 Super is positioned as a production-ready model optimized for enterprise deployment.
Nvidia Fans, Investors Gear Up For AI Conference
The development comes ahead of Nvidia’s GTC conference, kicking off in San Jose next Monday. The annual developer conference typically features the unveiling of new AI chips, data-center systems, and software tools, with this year’s edition expected to place a heavy focus on agentic AI platforms.
Earlier in the week, news leaked of a forthcoming agentic platform from Nvidia called NemoClaw, designed for enterprises to build AI agents to perform tasks for their own workforces.
Latest Retail’s View On NVDA
On Stocktwits, retail sentiment has dropped since the start of last week and was ‘bearish’ as of late Wednesday, amid ‘low’ message volume. Users discussed the new AI model, the Nebius investment, and shares, but more importantly, how the U.S.-Iran war would continue to weigh on the broader market.
“$NVDA Looking to be a bloodbath tomorrow. Sadly no solution for Iran target practicing on oil tankers in the straight of H, which means oil will keep climbing. Markets are beginning to realize prez’s promises won’t fix the problem. Let’s hope NVDA can hold strong!!” said a user.
Nvidia shares have risen for three straight sessions, gaining 4.6%, but remain marginally lower year to date.
For updates and corrections, email newsroom[at]stocktwits[dot]com.<