Nvidia Launches Advanced Open-Source AI Models Amid Rising Chinese Competition

SAN JOSE / BEIJING — Nvidia on Monday launched a new family of advanced open‑source artificial intelligence models—part of a broader strategic shift as competition from Chinese AI developers accelerates around the globe.

The lineup, branded the Nemotron 3 series, includes the immediate release of a compact Nemotron 3 Nano model, with larger, more powerful versions planned for early 2026. Designed to be more efficient, cost‑effective, and capable of handling complex language, reasoning, and multi‑step tasks, these models mark Nvidia’s biggest push yet into the open‑source model arena, traditionally dominated by start‑ups and specialized AI labs.

Nvidia’s move comes amid a boom in open‑source AI offerings from Chinese companies such as DeepSeek, Moonshot AI, and others that are rapidly gaining traction both domestically and internationally. Models from these firms are attracting developers and enterprises alike, in part due to their accessibility and competitive performance on tasks ranging from coding to large‑context reasoning.

The company has also bolstered its open‑source push by acquiring SchedMD, an AI software firm known for its open‑source workload scheduling technology, which will remain under open licenses even after integration into Nvidia’s toolstack. This acquisition aims to enhance Nvidia’s ecosystem, particularly for large‑scale AI model training and inference infrastructure.

While Nvidia historically has been best known for its AI chips powering major cloud and generative AI applications, the blueprints it released with Nemotron are intended to give developers greater transparency and customization—in contrast to some closed models being pursued by other Western firms. 

Industry observers say the timing reflects broader dynamics in the global AI race: Chinese AI labs are proliferating open‑source models that are not only low‑cost but also perform strongly in benchmarks and real‑world use cases, challenging traditional Western dominance in the sector. 

The development comes against the backdrop of geopolitical tensions over technology and supply chains. There are ongoing debates around export controls, tech self‑sufficiency, and the future role of open‑source AI tools in everything from research to commercial products. Nvidia’s push into open‑source modelling signals its intention to remain at the centre of that ecosystem, even as Chinese models continue to evolve and spread globally.