nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16
main boardmodel
NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16 is the full-precision release of NVIDIA Nemotron 3.5 Super VL, a vision-language model for reasoning, coding, tool use, document understanding, and video understanding. Mixture-of-experts, 124B total, 12B active (512 experts, 22 per token). Context: 1M tokens (config). License: openmdw-1.1.
- Org
- nvidia
- Released
- 2026-10-12
- Status
- main board
- License
- openmdw-1.1 (permissive)
- Parameters
- 124.5B total, 12.0B active
- Architecture
- MoE · 512 experts, 22/token · NemotronH_Omni_Reasoning_V3
- Context
- 1,048,576 tokens
- Task
- image-text-to-text · vision
- Signal
- 101 HF likes · 34316 downloads
Runs on
real quant files; need = weights + FP16 KV cache for 8K tokens + 1GB. A sizing calculation, not tested.
- Q4 ~98.5 GB: 1× DGX Spark (128GB), CPU only, 128GB RAM (slow)
- Q8 ~143.2 GB: 2× DGX Spark (256GB), Mac 192GB, Mac 256GB, CPU only, 256GB RAM (slow)
- FP16 ~261.2 GB: Mac 512GB
Links
- Hugging Face
- Ollama
- GGUF: ggml-org/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-GGUF
- quant: nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4
- quant: nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4-LiLiCorr
- Unsloth: unsloth/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B
- GGUF: unsloth/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-GGUF
- Unsloth: unsloth/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-NVFP4
First published here 2026-10-12 · updated 2026-10-12 · markdown · JSON