nemotron-3.5-super-vl-120b-a12b

nvidia/nemotron-3.5-super-vl-120b-a12b

NVIDIA Build Card

Nemotron 3.5 Super early-access reasoning model for agentic, coding, tool-calling, and multimodal workloads

Context Length

262,144 tokens

Input Pricing

$0.18 / 1M

Output Pricing

$0.18 / 1M

Recent Latency

No probes

Model Architecture & Limits

Canonical IDnvidia/nemotron-3.5-super-vl-120b-a12b

Publisher / Originatornvidia

Model FamilyNemotron

ArchitectureMamba-2–Transformer hybrid Mixture of Experts (MoE) with Multi-Token Prediction (MTP) (Nemotron Hybrid MoE with 512 routed experts per MoE layer, a C-RADIO vision encoder, and a vision-language connector)

Context Window Limit262,144 tokens

Max Output Tokens4,096 tokens

Endpoint Protocolopenai-chat

Capabilities & Modal Surfaces

Chat Conversation✓ Yes

Tool / Function Calling✓ Yes

Reasoning / Thinking✓ Yes

Structured Output (JSON)✓ Yes

Vision / Multimodal✓ Yes

Audio Processing✓ Yes

License & Restrictions

**GOVERNING TERMS**: The trial service is governed by the [NVIDIA API Trial Terms of Service](https://assets.ngc.nvidia.com/products/api-catalog/legal/NVIDIA%20API%20Trial%20Terms%20of%20Service.pdf). Use of the model is governed by the [OpenMDW License Agreement, version 1.1](https://github.com/Ope

Commercial and research use permitted under governing model and API terms.