Nemotron 3.5 Super VL
Parameters
120B
Context Length
1.05M
Modality
Multimodal
Architecture
-
KV Cache Rate
128 KB/tok
License
-
Release Date
15 Oct 2026
Knowledge Cutoff
-
API Pricing (per 1M)
Input: $0.20 · Output: $0.50
Hardware Sizing
Precision
Context
Workload Memory: 227.02 GB
FP16 · 8K context
Base Weights: 223.52 GB
KV Cache: 1.00 GB
Runtime & Graphs: 2.50 GB
Minimum Required Hardware
8x NVIDIA H100 80GB
227.0 / 576.0 GB usable · TP 8
Determine requirements for your specific workload:
Open in VRAM Calculator
Architecture Diagram
Evaluation Benchmarks
Rank
#169
| Benchmark | Score | Rank |
|---|---|---|
Reasoning | 0.79 | 53 |
Data Analysis | 0.65 | 55 |
Agent Arena | -14.94 | 61 |
General | 0.51 | 62 |
Coding | 0.66 | 62 |
LiveBench Average | 0.51 | 62 |
Web Development | 1399 | 82 |
General Text | 1412 | 110 |
Intelligence Index | 0.14 | 205 |
Rankings
Overall Rank
#169
Coding Rank
#119
About Nemotron 3.5 Super VL
Nemotron 3.5 Super VL is an open multimodal mixture-of-experts model from NVIDIA with 120B total and 12B active parameters. It processes text, images, video, documents, and GUI screenshots for advanced visual and agentic reasoning.
Technical Specifications
Resources
About Nemotron 3.5
NVIDIA's Nemotron 3.5 family of open models delivers leading efficiency and accuracy for agentic AI applications. Built on an advanced hybrid Mamba-2 SSM and sparse Mixture-of-Experts architecture with native Multi-Token Prediction (MTP) and speculative decoding harnesses, Nemotron 3.5 models are purpose-built for the high-volume execution layer of always-on agents.