NVIDIA推出Nemotron 3 Nano Omni模型，多模态推理效率提升9倍
Odaily星球日报讯 NVIDIA 在 X 平台发文表示，今日推出开源多模态模型 Nemotron 3 Nano Omni。该模型采用 30B-A3B 混合专家架构（MoE），支持 256K 上下文，可统一处理视频、音频、图像及文本输入。与同类交互水平的开源全能模型相比，该模型吞吐量提升达 9 倍，大幅降低推理成本并提高扩展性。目前 Nemotron 3 Nano Omni 已在 Hugging Face、OpenRouter 及 NVIDIA NIM 上线，并获得 Aible、Applied Scientific Intelligence、H Company 等企业采用。
Disclaimer: OKX Orbit content is provided for informational purposes only. Learn more
Replies
Related Flash News
Blockbeats•1s agoPreview: Nvidia's Q1 earnings report and the minutes of the Federal Reserve meeting will appear together tonight
Odaily•31m agoBitget UEX continues to expand its capacity and now supports more than 340 US stock assets
币界网•4h agoA trillion-parameter open-source model runs 981 words per second, and Cerebras tests KimiK2.6 speeds up 29 times
币界网•6h agoRuixun Group: Nvidia's earnings report has weakened the impact of AI investor sentiment
币界网•7h agoDuan Yongping made a massive purchase of Tesla, reduced his Apple holdings, and sold out Alibaba
TechFlow•14h agoNvidia's earnings battle is coming, and Bybit's three major trading products help users capture AI fluctuations across borders
Odaily•22h agoThe U.S. semiconductor sector staged a "V-shaped" reversal during the session, and the decline of many stocks narrowed sharply or even turned up
ChainCatcher•1d agoU.S. stock index futures edged lower, and semiconductor stocks continued to fall
币界网•1d agoBitget Wallet adds 130 new Xstocks tokenized stocks
币界网•1d agoZyphra releases the first MI355X inference review: long text is close to B200