Nvidia and Meta each released open-weight AI models this week, joining a push by more than 20 U.S. tech companies to compete with open models from Chinese labs. Nvidia's Nemotron 3.5 Lightning followed Meta's Muse Glimmer by a day, with both companies framing the releases as evidence that U.S. firms can match near-frontier open-source AI.
Nvidia released Nemotron 3.5 Lightning a day after Meta launched Muse Glimmer on Monday, extending a broader push by U.S. tech firms to counter open-weight AI models from Chinese labs.
Two releases in two days
Nemotron 3.5 Lightning builds on Nvidia's Nemotron 3 family, released in December. Nvidia says its models are truly open source because it publishes the training datasets, techniques, and model weights for developers to inspect. Meta's release stems from CEO Mark Zuckerberg's plan to open the weights for its latest model, Muse Spark 1.2. Both new models are smaller than the companies' flagship systems and are meant to run on laptops for on-device digital agents.
Competing with Chinese open models
The releases follow a July 24 open letter in which more than 20 U.S. tech companies urged policymakers not to place early restrictions on open-weight AI, including models built through distillation. Nvidia and Meta now face competition from Chinese labs such as Moonshot AI, DeepSeek, and Alibaba's Qwen. Box CEO Aaron Levie, a signatory of the letter, said the releases mark a turning point. According to Levie: "There's a very firm flag in the ground that America will have near-frontier open-source models".
Developers remain skeptical of Meta
Meta previously pursued open weights with Llama, but the April 2025 release of Llama 4 left developers unimpressed. The company then spent billions overhauling its AI unit and installed Scale AI CEO Alexandr Wang to lead it, and Wang's group has since released proprietary models under the Muse brand. Uniphore CEO Umesh Sachdev said Meta's shift from open to proprietary weights damaged trust with developers, who he said almost feel betrayed. Still, Sachdev said he is rooting for domestic AI companies because more competition drives down token costs. Forrester analyst Charlie Dai called Meta's move strategically important because it restores a major U.S. frontier AI vendor to the open ecosystem.
Source: CNBC
Trading involves risk.
Risk Disclaimer
