BenchAtlas

Rankings / NVIDIA

Nvidia Nemotron 3 Super 120B A12b

open weights

released 2026-03-11

BenchAtlas Index

as of 2026-07-14
40.8
thinking
rank #289
7 families · 4 categories · medium

Benchmark evidence

29 results
agentic coding
τ²-Benchthinking67.8 %independentArtificial Analysis
τ²-Bench
subset=banking
thinking10.1 %independentArtificial Analysis
coding
SciCodethinking36.0 %independentArtificial Analysis
external indices
AA Coding Indexthinking37.7 pointsindependentArtificial Analysis
Intelligence Index v4.1thinking25.4 pointsindependentArtificial Analysis
human preference
Coding (style control)older version
arena=text · category=coding · style_control=true
1410.3
1396.6–1424.0
communityLMArena Leaderboard Dataset
Industry Software And It Services (style control)older version
arena=text · category=industry_software_and_it_services · style_control=true
1404.3
1393.0–1415.6
communityLMArena Leaderboard Dataset
Hard Prompts English (style control)older version
arena=text · category=hard_prompts_english · style_control=true
1395.4
1382.5–1408.4
communityLMArena Leaderboard Dataset
Expert (style control)older version
arena=text · category=expert · style_control=true
1395.3
1371.6–1419.1
communityLMArena Leaderboard Dataset
English (style control)older version
arena=text · category=english · style_control=true
1384.3
1374.4–1394.2
communityLMArena Leaderboard Dataset
Hard Prompts (style control)older version
arena=text · category=hard_prompts · style_control=true
1380.5
1371.1–1389.8
communityLMArena Leaderboard Dataset
Math (style control)older version
arena=text · category=math · style_control=true
1375.0
1349.7–1400.2
communityLMArena Leaderboard Dataset
Industry Legal And Government (style control)older version
arena=text · category=industry_legal_and_government · style_control=true
1368.3
1342.5–1394.1
communityLMArena Leaderboard Dataset
Industry Life And Physical And Social Science (style control)older version
arena=text · category=industry_life_and_physical_and_social_science · style_control=true
1366.2
1349.0–1383.4
communityLMArena Leaderboard Dataset
Overall (style control)
arena=text · category=overall · style_control=true
1361.4
1354.1–1368.7
communityLMArena Leaderboard Dataset
Longer Query (style control)older version
arena=text · category=longer_query · style_control=true
1360.6
1347.9–1373.4
communityLMArena Leaderboard Dataset
Industry Business And Management And Financial Operations (style control)older version
arena=text · category=industry_business_and_management_and_financial_operations · style_control=true
1350.1
1334.7–1365.6
communityLMArena Leaderboard Dataset
Multi Turn (style control)older version
arena=text · category=multi_turn · style_control=true
1349.4
1333.2–1365.7
communityLMArena Leaderboard Dataset
Instruction Following (style control)older version
arena=text · category=instruction_following · style_control=true
1345.7
1332.7–1358.7
communityLMArena Leaderboard Dataset
Non English (style control)older version
arena=text · category=non_english · style_control=true
1336.5
1327.1–1345.8
communityLMArena Leaderboard Dataset
Russian (style control)older version
arena=text · category=russian · style_control=true
1328.4
1307.8–1349.0
communityLMArena Leaderboard Dataset
Industry Writing And Literature And Language (style control)older version
arena=text · category=industry_writing_and_literature_and_language · style_control=true
1326.3
1311.8–1340.8
communityLMArena Leaderboard Dataset
Exclude Ties (style control)older version
arena=text · category=exclude_ties · style_control=true
1324.0
1313.7–1334.4
communityLMArena Leaderboard Dataset
Industry Entertainment And Sports And Media (style control)older version
arena=text · category=industry_entertainment_and_sports_and_media · style_control=true
1316.8
1299.7–1333.8
communityLMArena Leaderboard Dataset
Creative Writing (style control)older version
arena=text · category=creative_writing · style_control=true
1303.4
1285.1–1321.8
communityLMArena Leaderboard Dataset
knowledge science
GPQA Diamond
implementation=artificial-analysis
thinking80.0 %independentArtificial Analysis
Humanity's Last Exam
implementation=artificial-analysis
thinking19.2 %independentArtificial Analysis
long context instruction
AA-LCRthinking60.0 %independentArtificial Analysis
IFBenchthinking71.5 %independentArtificial Analysis

Agent + model results

systems, not bare-model scores
agent + model Artificial Analysis harness + Nvidia Nemotron 3 Super 120B A12bTerminal-Bench 2.138.6 %independentArtificial Analysis
agent + model Artificial Analysis harness + Nvidia Nemotron 3 Super 120B A12bTerminal-Bench Hard28.8 %independentArtificial Analysis

These scores measure the whole agent system (scaffold, tools, budgets) — they are never merged into the bare model’s numbers.