[whatmodel.sh](../../README.md) / Nemotron 3.5 Super VL 120B A12B # Nemotron 3.5 Super VL 120B A12B NVIDIA · released 2026-10 · 120B params (12B active) · open-mdw-1.1 **score 67.2** ±1.0 · #110 of 732 · from 10 benchmark results across 2 sources Score within the [LMArena (text)](https://whatmodel.sh/?set=lmarena-text&score=set) comparison set only. ## What it's good at ● marks Nemotron 3.5 Super VL 120B A12B among every model on that task or test, worst to best. 
task                vs the others          rank  counts  tests
agent tasks         ▁▁▃▆█▄▅●▄▄▂▁     #98 of 313     48%      7
coding              ▁▁▁▃▅▇█▇●▄▃▁     #94 of 469     32%      6
chat & writing      ▁▁▁▃▆▄▆█▆●▇▃   #103 of 402*     11%      1
vision & documents  ▁▁▂▂▁▃▄▂●▇█▄    #78 of 153*    7.7%      1
web search          █··········●       #1 of 2*    3.0%      1
agent tasks · #98 of 313 · 48% of its score Beats [Grok 4.20 (0309 reasoning)](grok-4-20.md), [GPT-5.1](gpt-5-1.md), [o3-pro](o3-pro.md) here, though they rank higher overall. Trails [Mistral Medium 3.5 (128B)](mistral-medium-3-5.md), [Hy3 preview (295B-A21B)](hy3-preview.md), [MiniMax-M2.7](minimax-m2-7.md) here, though they rank lower overall. 
test                                     result  vs the others in its run   difficulty            counts  why
Arena Agent (October 2026 aggrega…  -0.1494 IPS  ▁●▁▃▃▂█▁▂▂▁▁ beats 3/59    avg -0.379 IPS           29%  sharp
SWE-bench Verified (maker)                71.1%  ▁▁▁▁▁▁▂▄●▆▁▁ beats 87/138  hard · avg 22%          8.7%  sharp · at its level · small run
Terminal-Bench 2.1 (maker)                60.7%  ▁·▁·▁▁▄▃●█▅█ beats 27/63   hard · avg 16%          5.2%  sharp · at its level · small run
BrowseComp (maker)                        78.0%  ▂·▃▁▂▂▆▅█●▂▄ beats 37/46   hard · avg 31%          3.0%  at its level · small run
τ³-bench Banking (maker)                  18.9%  ▂▂·▅█▂▃▃▃●▂▅ beats 17/24   very hard · avg 6.5%    1.6%  blunt · small run
SWE-bench Multilingual (maker)            67.7%  ▁··▁▃▁▅●█▂▂▁ beats 34/65   hard · avg 18%             —  not counted yet: no other model in the …
SWE-bench Pro (split unspecified)…        51.8%  ▂·▁▁▁▁▂▃●▇▂▁ beats 30/58   very hard · avg 3.9%       —  not counted yet: no other model in the …
coding · #94 of 469 · 32% of its score Beats [Grok 4.20 (0309 reasoning)](grok-4-20.md), [GPT-5.4 mini](gpt-5-4-mini.md), [GPT-5.1](gpt-5-1.md) here, though they rank higher overall. Trails [Llama 3.1 Nemotron Ultra 253B v1](llama-3-1-nemotron-ultra-253b.md), [Qwen3 235B-A22B Thinking 2507](qwen3-235b-a22b-thinking-2507.md), [o4-mini](o4-mini.md) here, though they rank lower overall. 
test                                result  vs the others in its run   difficulty            counts  why
Arena WebDev (Code Arena)           1399.2  ▁▁▃▂▅●▃▆▅▅▁▂ beats 45/129  avg 981                  20%  sharp
SWE-bench Verified (maker)           71.1%  ▁▁▁▁▁▁▂▄●▆▁▁ beats 87/138  hard · avg 22%          8.7%  sharp · at its level · small run
SciCode subproblems (maker)          47.1%  ▃··▃▃▃·▃●█·▅ beats 5/10    hard · avg 21%          3.1%  at its level · small run
LiveCodeBench v6 (window unspecif…   88.2%  ▂▃▄▄▅██▆▄▇●▆ beats 57/67   medium · avg 38%           —  not counted yet: no other model in the …
SWE-bench Multilingual (maker)       67.7%  ▁··▁▃▁▅●█▂▂▁ beats 34/65   hard · avg 18%             —  not counted yet: no other model in the …
SWE-bench Pro (split unspecified)…   51.8%  ▂·▁▁▁▁▂▃●▇▂▁ beats 30/58   very hard · avg 3.9%       —  not counted yet: no other model in the …
chat & writing · #103 of 402 · 11% of its score Beats [LongCat 2.0 (1.6T-A48B)](longcat-2-0.md), [Inkling Small (266B-A12B)](inkling-small.md), [GPT-5.4 nano](gpt-5-4-nano.md) here, though they rank higher overall. Trails [Mistral Large 3 (675B-A41B)](mistral-large-3.md), [Gemma 4 12B](gemma-4-12b.md), [Kimi K2 Instruct (0905)](kimi-k2-0905.md) here, though they rank lower overall. 
test            result  vs the others in its run    difficulty  counts  why
LMArena (text)  1412.3  ▁▁▁▂▃▃▃▆▆●█▃ beats 259/385  avg 1260       11%  blunt
vision & documents · #78 of 153 · 7.7% of its score Beats [DeepSeek V4 Flash Vision (Exp)](deepseek-v4-flash-vision-exp.md), [GPT-5.4 nano](gpt-5-4-nano.md) here, though they rank higher overall. Trails [Gemma 4 12B](gemma-4-12b.md), [Qwen3.5 9B](qwen3-5-9b.md), [GPT-4.1](gpt-4-1.md) here, though they rank lower overall. 
test                             result  vs the others in its run   difficulty  counts  why
Arena vision (style controlled)  1204.0  ▁▁▂▂▂▂▃▄▅●█▇ beats 76/152  avg 1061      7.7%  blunt
web search · #1 of 2 · 3.0% of its score Beats [Nemotron 3.5 Bolt 74B A9B](nemotron-3-5-bolt-74b-a9b.md) here, though they rank higher overall. 
test                result  vs the others in its run  difficulty      counts  why
BrowseComp (maker)   78.0%  ▂·▃▁▂▂▆▅█●▂▄ beats 37/46  hard · avg 31%    3.0%  at its level · small run
other tests (4) 
test                                result  vs the others in its run    difficulty        counts  why
GPQA Diamond (maker)                 89.5%  ▁▁▁▁▂▄▅▅██●▄ beats 272/322  hard · avg 47%      9.1%  sharp · small run
IFBench (aggregation unspecified)…   70.9%  ▂▆▂▄▄▄▂▄▄●█▇ beats 58/86    medium · avg 44%       —  not counted yet: no other model in the …
MMLU-ProX (maker)                    76.9%  ▁·▁▁▁▂▁▃●█▅▄ beats 15/36    medium · avg 58%       —  not counted yet: no other model in the …
MMLongBench-Doc (maker)              62.8%  ▃···▃···▃▅█● beats 9/9                             —  not counted yet: no other model in the …
## How the score was worked out Each result implies a score on its own, after adjusting for how hard that test is and who ran it. The model's score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its *counts* column above; here it is added up by who ran the tests. [Method](../method.md) 
source                                    results  share of the score
LMArena leaderboard dataset                     4   69% ▇▇▇▇▇▇▇▇
NVIDIA vendor-reported benchmark scores…        6   31% ▇▇▇▇
prior: pull toward the middle                        0% ▇
weighted average                                   = 67.2
Licences: LMArena leaderboard dataset ([CC BY 4.0](https://creativecommons.org/licenses/by/4.0/))
Not counted (6) 
source                  test                            score  why
NVIDIA vendor-reporte…  MMLU-ProX                       76.9%  too few models have it yet
NVIDIA vendor-reporte…  LiveCodeBench v6 (window unsp…  88.2%  too few models have it yet
NVIDIA vendor-reporte…  IFBench (aggregation unspecif…  70.9%  too few models have it yet
NVIDIA vendor-reporte…  SWE-bench Multilingual          67.7%  too few models have it yet
NVIDIA vendor-reporte…  SWE-bench Pro (split unspecif…  51.8%  too few models have it yet
NVIDIA vendor-reporte…  MMLongBench-Doc                 62.8%  too few models have it yet
## How to run it ### Other ways to run it 1. Run it as published (BF16, needs ~240 GB, NVIDIA only, from [huggingface.co/nvidia](https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16/blob/1f8953f7f0935358e3605150d22bc15df062061c/README.md)) ```sh vllm serve nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16 \ --served-model-name nemotron-3.5-super-vl \ --tensor-parallel-size 4 \ --trust-remote-code \ --gpu-memory-utilization=0.85 \ --allowed-local-media-path=/ \ --max-num-seqs=32 \ --limit-mm-per-prompt '{"video": 1, "image": 1}' \ --video-pruning-rate 0.5 \ --media-io-kwargs '{"video": {"fps": 2, "num_frames": 256}}' \ --enable-auto-tool-choice \ --tool-call-parser qwen3_coder \ --reasoning-parser nemotron_v3 \ --speculative_config.method mtp \ --speculative_config.num_speculative_tokens 3 ``` Model repository and documentation: [huggingface.co/nvidia](https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16) [interactive version](https://whatmodel.sh/model/nemotron-3-5-super-vl-120b-a12b) · [JSON](https://whatmodel.sh/api?format=json&all=1)