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
Odaily•10h agoThe U.S. chip and semiconductor sector fell collectively, with the Nasdaq Composite Index pulling back below 26,000 points
Odaily•11h agoNvidia's stock price continued to hit a record high
TechFlow•1d agoBybit TradFi now supports 300+ popular US stock trades
ChainCatcher•1d agoThe three major U.S. stock indexes closed slightly higher, and large chip stocks hit new highs
Odaily•1d agoMusk will participate in Trump's trip to China as a member of the U.S. delegation
ChainCatcher•1d agoU.S. stocks opened down, the Dow fell 0.12%, and AI chip stocks diverged
ChainCatcher•1d agoThe three major indexes fluctuated slightly in the U.S. stock market, with Tesla (TSLA) falling 0.98%
Odaily•1d agoHuang Renxun: The 60-year computing paradigm is over, and people who can use AI will have an advantage
TechFlow•1d agoU.S. stocks opened, and crypto stocks fell
TechFlow•1d agoJPMorgan: Emerging markets offer investors a lower-cost way to invest in AI