Meta released Muse Glimmer, a 30-billion-parameter model under Apache 2.0 that runs locally on consumer hardware with 24GB+ VRAM, distilled from its closed flagship Muse Spark 1.2. Meta also plans to release open-weight Muse Spark 1.2 soon, marking its return to open weights after more than a year.
Nvidia open-sourced Nemotron 3.5 Lightning, a lightweight model companies can download, modify and use without permission or payment. The company is also developing Nemotron 4 with at least 1 trillion parameters to rival leading open-source models globally, though no release date has been set.
Alibaba's Qwen3.8-Max scored 56 on the Artificial Analysis Intelligence Index, a 10-point jump over Qwen3.7 Max, beating Claude Opus 4.8 and splitting with GPT-5.6 Sol. The model is free, open-weight, and priced at $2 per million tokens.
Meta CEO Mark Zuckerberg published a 6,500-word essay warning that concentrated AI control by a few companies is the biggest AI risk, arguing for lower US barriers on open-source AI to compete with Chinese rivals. The essay coincided with Meta's Muse Glimmer release and a planned $1 billion data center investment.
Nvidia signed memorandums of understanding with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs and KKR to finance AI infrastructure projects. The $500 billion fund could impact enterprise chip pricing and availability, with analysts warning it may exacerbate the shortage of AI chips for smaller buyers.
A SaferAI evaluation found Z.ai's open-weight GLM-5.2 is only months behind GPT-5.5 and Claude Opus 4.7 on cyber and bio capability benchmarks, but with fewer safety mitigations. The finding highlights a growing gap as open-weight models catch up on capability but not on safety.
Unsloth AI released the first desktop app to train and run AI models locally, combining training, inference, image/video diffusion, fine-tuning and agents into one fully open-source application. NVIDIA is bringing support for Unsloth Desktop alongside new open models and a 2x performance boost on its DGX and RTX systems.