NVIDIA releases Nemotron 3.5 Super VL: a 120B multimodal model built for AI agents
NVIDIA launched Nemotron 3.5 Super VL, a 120B multimodal model for agents. It activates 12B params per request and reasons across text, images, documents, and video while calling tools.
NVIDIA's new Nemotron 3.5 Super VL is a 120-billion-parameter multimodal model built for agents, with only 12 billion parameters active per request.
What happened
NVIDIA released Nemotron 3.5 Super VL, a large multimodal model built for AI agents, announced Friday, October 16, 2026. The model has 120 billion total parameters but activates only 12 billion for each request. It can reason across text, images, documents, and video while calling tools. NVIDIA says it can deliver up to five times more throughput than comparable open models on a single GPU, though customer-reported improvements vary by deployment.
Why it matters
The release lands in the middle of the personal-agent race, one day after Meta shipped the 816B open-weight Muse Spark Open. Sparse-activation multimodal models like Nemotron 3.5 Super VL push agent-capable models toward cheaper single-GPU deployment, which matters for builders deciding what runs their agents.
Sources
- The Neuron: Everything that happened in AI today, Friday, October 16, 2026 — https://www.theneuron.ai/digest/everything-that-happened-in-ai-today-friday-october-16-2026/
- NVIDIA Blog — https://blogs.nvidia.com/
Verified Oct 16, 2026.
Sources
Get updates like this every morning
- ② Card on Stripe
- ③ 7 days free
Then $2/month · cancel anytime in one click