Magnitude
NVIDIA logo

Nemotron 3.5 Super VL 120B-A12B

by NVIDIA · Released · OpenMDW 1.1

Workstation-class hybrid MoE vision model for reasoning, coding, and agentic workflows.

Parameters
120B · 12B active
Context
1M tokens
Input
Text + image
Speculative decoding
LiLiCorr
Intelligence
46%
License
OpenMDW 1.1
Variants
NVFP4 QAT · Q4 · Q5 · Q6 · Q8
Source
Hugging Face ↗

Downloads

Sizes include the vision projector and speculative draft Magnitude installs. Every file is pinned to an immutable Hugging Face commit.

Variant Format Download Fidelity
NVFP4 QAT NVFP4QAT 80.8 GB 60/100
Q4 UD-Q4_K_XL 90.0 GB 40/100
Q5 UD-Q5_K_XL 114 GB 50/100
Q6 UD-Q6_K_XL 124 GB 60/100
Q8 Q8_0 135 GB 80/100

How it compares

Intelligence across every model in the catalog: each model's Artificial Analysis Intelligence Index score as a percentage of the top model on v4.3.2, as of Sep 29, 2026.

  1. Qwen3.8 Flash Next 69%
  2. Qwen3.8 27B 59%
  3. Nemotron 3.5 Super VL 120B-A12B 46%
  4. Gemma 4 31B 33%
  5. MiniCPM-V 4.7 35B-A3B 33%
  6. Nemotron 3.5 Bolt 74B-A9B 33%
  7. Qwen3.6 35B-A3B 32%
  8. Muse Glimmer 30B 30%
  9. Gemma 4 26B-A4B 29%
  10. Liquid LFM2.5 24B-A2B 27%
  11. Gemma 4 12B 25%
  12. Liquid LFM3 1.7B 24%
  13. Qwen3.5 4B 23%
  14. Liquid LFM3-VL 2B 22%
  15. MiniCPM5 2B 22%
  16. Nemotron 3.5 Lightning 30B-A3B 22%
  17. Qwen3.5 9B 19%
  18. Gemma 4 E4B 15%
  19. Liquid LFM2.5 2.6B 15%
  20. MiniCPM5 1B 15%
  21. Gemma 4 E2B 14%
  22. Liquid LFM2.5 8B-A1B 13%
  23. MiniCPM-V 4.7 1B 10%

Works with your agent

One click in Magnitude connects Nemotron 3.5 Super VL 120B-A12B to any of these harnesses.

  • Pi
  • OpenCode
  • Hermes
  • OpenClaw
  • Codex
  • Claude Code
  • Oh My Pi
  • Cline

Related models

FAQ

How much memory does Nemotron 3.5 Super VL 120B-A12B need?

The smallest download is 80.8 GB and the largest is 135 GB. Plan for the weights plus a few gigabytes for context. Magnitude measures your free memory and only recommends variants that fit with room to work.

Can I run Nemotron 3.5 Super VL 120B-A12B on a Mac?

Yes. Magnitude runs GGUF models on Apple Silicon through Metal, and on NVIDIA, AMD, and CPU-only machines on Windows and Linux. The app estimates tokens per second for your exact chip before you download.

Which variant should I pick?

Magnitude picks for you: the highest-fidelity variant that fits your memory. If you choose manually, Q8 keeps the most fidelity, and NVFP4 QAT is the smallest at 80.8 GB.

Catalog data from magnitudedev/magnitude, updated Oct 28, 2026.

Run Nemotron 3.5 Super VL 120B-A12B on your machine

Download Magnitude