Nemotron 3 Ultra vs Qwen3.7-Max

Jensen Huang announced at GTC Taipei 2026 that the Agentic AI era has arrived, shifting AI from content generation to autonomous task execution. NVIDIA launched infrastructure products like Vera Rubin and Vera CPU, driving a computing paradigm shift where AI becomes a direct generator of profit and GDP.

NVIDIA today launches Nemotron 3 Ultra, a 550B-parameter open model built on the same architecture as Nemotron 3 Super, optimized for long-running AI agents. It employs LatentMoE to quadruple the number of experts at the same inference cost, introduces multi-token prediction to boost single-user inference speed, and is released under the Linux Foundation’s Open MDW license to enable enterprise deployment.

NVIDIA introduces the 550B-parameter Neotron 3 Ultra, a mixture-of-experts agent model trained for task orchestration, outperforming many trillion-parameter open agents on benchmarks, with full data and recipe transparency to enable enterprise on-prem deployment and fine-tuning.

2026年前沿模型训练方案转向多专家蒸馏（MOPD），通过多个领域专家模型训练后合并为一个通用模型，显著提升效率与性能。

NVIDIA 的 NeMoTron 3 Ultra 是一款开源 AI 模型，速度快但编码能力有限，适合非复杂任务。

Nemotron 3 Ultra is NVIDIA's latest open-source sparse mixture of experts model with 550 billion total parameters and 55 billion active parameters, designed for agentic use cases; the article demonstrates steps to build an autonomous research agent with NemoHermes, including configuring an NVIDIA API key to use build.nvidia.com endpoints, selecting Ultra as the inference model, and completing a simplified auto-research task (e.g., optimizing NanoChat) to produce real optimization results.

NVIDIA released the powerful open-source MoE model Nemotron 3 Ultra with 550B parameters and a 1M context window, designed for long-running agent workloads, offering up to 5x performance improvement and 30% cost reduction. Meanwhile, Anthropic published research indicating early signs of Recursive Self-Improvement (RSI) in AI, revealing Claude now contributes over 80% of Anthropic's code.

NVIDIA released Nemotron 3 Ultra, a 550B MoE open model for long-running agents, offering 5x faster inference and 30% lower costs.

The U.S. government's new executive order balances AI development with security, Qwen3.7-Max enters the top three models, and AI vulnerability detection breaks copyright alignment.

ITBench-AA is a new benchmark series evaluating models on agentic enterprise IT tasks, starting with Site Reliability Engineering tasks where frontier models score below 50% on ITBench-AA's SRE tasks benchmark model performance on Kubernetes incident response, where models and agents must diagnose live systems by reading logs, tracing dependencies, and identifying root-cause entities across complex infrastructure.

Qwen3.7-Max has become the second-best AI programming model globally, scoring 1541 on Code Arena, trailing only Claude. Designed for production use, it can handle 35-hour tasks, over 1,000 tool calls, and complete two-week projects in hours.

阿里巴巴推出全新升级的超大规模语言模型 Qwen3.7-Max，该模型专为代理中心工作设计，如编码、办公和生产任务以及长期自主执行。相较于前代 Qwen3.6，Qwen3.7-Max 在编码和代理基准测试中取得了显著进步，并引入了显式提示缓存功能，以优化重复上下文的处理。

阿里巴巴推出Qwen3.7-Max，作为面向代理时代的最新旗舰模型，它是一个多功能的基础模型，适用于能够实际完成任务的代理。该模型在编码代理方面表现出色，能够进行前端原型设计、多文件重构和实际调试。此外，它还是一个可靠的办公和生产力助手。

Qwen3.7-Max在编码代理和通用代理的基准测试中表现出色，尤其在最难的推理基准上表现出色，并在通用能力和多语言支持方面脱颖而出。

Qwen3.7-Max 在人工智能分析指数上获得了56.6分，比Qwen3.6-Max-Preview提高了4.8分。它在科学推理、代理能力、编码能力和减少幻觉方面都有显著提升。

Alibaba Cloud's Qwen3.7-Max scores 56.6 to rank 5th globally and 1st domestically in Artificial Analysis benchmark, soon available via Alibaba Cloud's BaiLian API.