NVIDIA

Nemotron 3 Super 120B A12B

Nemotron 3 Super 120B A12B is an open-weights 120B-parameter model by NVIDIA. It holds an Intelligence score of 64.6 on the local.ai leaderboard (2026-08 snapshot).

MoEopen weights
64.6
local.ai Intelligence
Parameters
120B
MoE · 12B active parameters
Benchmarks (local.ai snapshot)
τ²-bench
68%
GAIA
70%
GDPval
60%
HuggingFace metadata · updated 2026-04-29
repo
nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-BF16
license
other
downloads
849,022
likes
418

Metadata refreshed from the HuggingFace API (snapshot 2026-08-15T15:53:42.405Z). Context length and long-form descriptions are the next pipeline stage.

Does it fit your hardware?

Detect or select your GPU first. Green = weights + context headroom, yellow = tight, grey = does not fit.

Q4_K_M 69.6GQ6_K 108GQ8_0 132GFP16 240G
Quantization≈ File sizeNote
IQ2_XXS32.4 GBExtreme compression, largest quality loss
IQ3_XXS43.2 GBVery small
Q3_K_M55.2 GBSmall
NVFP460 GBNVIDIA 4-bit
Q4_K_S64.8 GBCommon
Q4_K_M69.6 GBMost common sweet spot
Q5_K_M91.2 GBBalanced
Q6_K108 GBNear-lossless
Q8_0132 GBNear-lossless
FP16240 GBOriginal weights

Actual files on HuggingFace

FileSize
model-00049-of-00050.safetensors4.66 GB
model-00006-of-00050.safetensors4.65 GB
model-00001-of-00050.safetensors4.65 GB
model-00012-of-00050.safetensors4.65 GB
model-00018-of-00050.safetensors4.65 GB
model-00024-of-00050.safetensors4.65 GB
GGUF: unsloth/NVIDIA-Nemotron-3-Super-120B-A12B-GGUF · 22,619 downloads
FileSize
BF16/NVIDIA-Nemotron-3-Super-120B-A12B-BF16-00001-of-00005.gguf45.82 GBdownload
BF16/NVIDIA-Nemotron-3-Super-120B-A12B-BF16-00002-of-00005.gguf44.61 GBdownload
BF16/NVIDIA-Nemotron-3-Super-120B-A12B-BF16-00003-of-00005.gguf44.55 GBdownload
BF16/NVIDIA-Nemotron-3-Super-120B-A12B-BF16-00004-of-00005.gguf44.61 GBdownload
BF16/NVIDIA-Nemotron-3-Super-120B-A12B-BF16-00005-of-00005.gguf45.34 GBdownload
MXFP4_MOE/NVIDIA-Nemotron-3-Super-120B-A12B-MXFP4_MOE-00001-of-00003.gguf0.01 GBdownload
MXFP4_MOE/NVIDIA-Nemotron-3-Super-120B-A12B-MXFP4_MOE-00002-of-00003.gguf46.41 GBdownload
MXFP4_MOE/NVIDIA-Nemotron-3-Super-120B-A12B-MXFP4_MOE-00003-of-00003.gguf30.01 GBdownload
Q8_0/NVIDIA-Nemotron-3-Super-120B-A12B-Q8_0-00001-of-00004.gguf0.01 GBdownload
Q8_0/NVIDIA-Nemotron-3-Super-120B-A12B-Q8_0-00002-of-00004.gguf45.64 GBdownload
Q8_0/NVIDIA-Nemotron-3-Super-120B-A12B-Q8_0-00003-of-00004.gguf45.90 GBdownload
Q8_0/NVIDIA-Nemotron-3-Super-120B-A12B-Q8_0-00004-of-00004.gguf28.10 GBdownload
GGUF: lmstudio-community/NVIDIA-Nemotron-3-Super-120B-A12B-GGUF · 8,943 downloads
FileSize
NVIDIA-Nemotron-3-Super-120B-A12B-Q4_K_M-00001-of-00003.gguf36.51 GBdownload
NVIDIA-Nemotron-3-Super-120B-A12B-Q4_K_M-00002-of-00003.gguf36.99 GBdownload
NVIDIA-Nemotron-3-Super-120B-A12B-Q4_K_M-00003-of-00003.gguf6.64 GBdownload
NVIDIA-Nemotron-3-Super-120B-A12B-Q6_K-00001-of-00003.gguf36.26 GBdownload
NVIDIA-Nemotron-3-Super-120B-A12B-Q6_K-00002-of-00003.gguf36.52 GBdownload
NVIDIA-Nemotron-3-Super-120B-A12B-Q6_K-00003-of-00003.gguf32.38 GBdownload
NVIDIA-Nemotron-3-Super-120B-A12B-Q8_0-00001-of-00004.gguf36.77 GBdownload
NVIDIA-Nemotron-3-Super-120B-A12B-Q8_0-00002-of-00004.gguf36.99 GBdownload
NVIDIA-Nemotron-3-Super-120B-A12B-Q8_0-00003-of-00004.gguf37.12 GBdownload
NVIDIA-Nemotron-3-Super-120B-A12B-Q8_0-00004-of-00004.gguf8.76 GBdownload

What you save self-hosting this model

Uses the model's own API price when it exists; otherwise the closest commercial equivalent.

Waiting for hardware detection…

RunLocal presents third-party leaderboard data (local.ai / Exo Labs, 2026-08) with its own fit and cost estimates. Verify pricing, licenses and file sizes at the source before deploying a model.