benchgap
NVIDIA · model

Nemotron 3 Ultra benchmark scores

As of 2026-11-21, Nemotron 3 Ultra (NVIDIA) has measured scores on 40 benchmarks and estimated scores on 121 more.

BenchmarkScoreSource
AA-GPQA Diamond86.7%measured
AA-HLE28.4%measured
AA-SciCode40.3%measured
Artificial Analysis Intelligence Index11.8%measured
CritPt3.1%measured
GDPval-AA33.1%measured
GPQA Diamond87.0%measured
HLE26.7%measured
IFBench81.7%measured
AA-LCR67.0%measured
LiveCodeBench v689.0%measured
AA-Omniscience Accuracy21.6%measured
SciCode44.6%measured
Terminal-Bench 2.156.4%measured
BrowseComp44.4%measured
HLE w/ tools37.4%measured
IFEval93.2%estimated ± 2.8 pp, medium confidence
LongBench v261.9%measured
VITA-Bench12.7%estimated ± 6.7 pp, low confidence
τ²-bench results83.3%measured
AA Coding Index49.3%measured
APEX-Agents25.2%estimated ± 9.4 pp, low confidence
APEX-Agents-AA7.4%estimated ± 4.4 pp, medium confidence
BioMysteryBench (human-difficult)31.4%estimated ± 6.8 pp, low confidence
DeepSearchQA85.4%estimated ± 4.5 pp, high confidence
SWE-bench Verified71.9%measured
AutomationBench48.6%estimated ± 9.1 pp, low confidence
BFCL v474.9%estimated ± 2.7 pp, medium confidence
CyberGym62.2%estimated ± 5.3 pp, low confidence
JobBench32.3%estimated ± 8.7 pp, medium confidence
SWE-bench Pro49.5%estimated ± 4.5 pp, high confidence
τ³-bench results70.9%measured
WideResearch55.9%estimated ± 3.5 pp, medium confidence
AA Agentic Index21.7%measured
MMLU-Pro86.8%measured
AA-IFBench81.4%measured
BBH99.8%estimated ± 2.8 pp, medium confidence
FailureSensorIQ66.5%estimated ± 5.0 pp, medium confidence
GPQA87.0%measured
MMLU89.9%estimated ± 0.9 pp, high confidence
Vals GPQA Diamond86.1%measured
Vals LiveCodeBench86.0%measured
Vals MMLU-Pro85.8%measured
MMMLU87.7%estimated ± 0.7 pp, medium confidence
Chess Puzzles36.1%estimated ± 4.7 pp, medium confidence
AA MMLU-Pro84.6%estimated ± 0.7 pp, medium confidence
Gert Labs26.8%estimated ± 7.1 pp, low confidence
AA-AnalystAgent44.8%estimated ± 5.9 pp, low confidence
AA EnterpriseOps-Gym35.8%estimated ± 7.2 pp, low confidence
AA Harvey LAB v1.081.7%measured
AA Tau3 Banking47.1%estimated ± 2.4 pp, low confidence
AA Terminal-Bench 2.179.8%estimated ± 2.5 pp, medium confidence
ApprenticeBench13.1%estimated ± 12.5 pp, low confidence
ARC-AGI-184.4%estimated ± 2.5 pp, medium confidence
ARC-AGI-240.6%estimated ± 9.7 pp, low confidence
CursorBench 3.147.9%estimated ± 4.8 pp, low confidence
CursorBench 3.234.2%estimated ± 4.4 pp, medium confidence
EBR-bench8.9%estimated ± 10.4 pp, low confidence
Terminal-Bench 3.02.2%estimated ± 6.9 pp, low confidence
FrontierCode 1.1 Main13.2%estimated ± 4.2 pp, low confidence
FrontierSWE v26.1%estimated ± 9.5 pp, low confidence
Furniture Assembly33.0%estimated ± 9.3 pp, medium confidence
MaxxBench34.6%estimated ± 5.7 pp, low confidence
MazeBench0.0%measured
Mystery Game Puzzles20.0%measured
NYT Connections62.6%estimated ± 13.8 pp, low confidence
OSWorld-Verified50.3%estimated ± 5.7 pp, low confidence
Terminal-Bench 4.10.0%estimated ± 10.9 pp, low confidence
Terminal-Bench Hard45.1%estimated ± 3.9 pp, medium confidence
Vals SWE-bench69.0%measured
Terminal-Bench 2.1 (Vals AI)50.9%measured
VulcanBench v384.3%estimated ± 5.6 pp, low confidence
CWE-Bench-AA27.5%measured
AA Cyber Index13.1%measured
DeepsecBench-AA5.8%estimated ± 8.9 pp, low confidence
HealthBench Professional (length-adjusted, Artificial Analysis)49.4%estimated ± 4.3 pp, low confidence
HealthBench Professional (raw, Artificial Analysis)32.3%estimated ± 0.8 pp, low confidence
AA ITBench39.8%estimated ± 4.0 pp, high confidence
Bug Hunt Bench14.8%estimated ± 10.5 pp, low confidence
CursorBench 4.030.8%estimated ± 3.4 pp, medium confidence
CWE-bench v157.4%estimated ± 5.1 pp, medium confidence
DeepSWE56.3%estimated ± 7.2 pp, low confidence
DRACO86.9%estimated ± 0.7 pp, low confidence
FrontierCode 2.026.1%estimated ± 9.3 pp, low confidence
FrontierPhysics1.0%estimated ± 4.7 pp, medium confidence
HLE w/o tools26.7%measured
OSWorld 2.01.8%estimated ± 14.7 pp, low confidence
OSWorld 2.157.3%estimated ± 5.3 pp, low confidence
OSWorld 2.1 Offline11.7%estimated ± 1.9 pp, low confidence
PostTrainBench v1.114.5%estimated ± 3.7 pp, medium confidence
SpireClimb0.0%estimated ± 9.3 pp, medium confidence
StealthMark16.7%measured
SWE Multilingual67.7%measured
Toolathlon-Verified50.4%estimated ± 2.3 pp, low confidence
VulcanBench Frontier v470.1%estimated ± 6.8 pp, low confidence
AA AutomationBench41.9%estimated ± 7.9 pp, low confidence
GDP.pdf17.0%estimated ± 4.6 pp, medium confidence
AA Terminal-Bench 4.04.3%estimated ± 11.0 pp, low confidence
ARC-AGI-30.3%estimated ± 7.2 pp, low confidence
ARC-AGI-3 Provider Adapter2.7%estimated ± 7.3 pp, low confidence
CWE-bench53.9%estimated ± 7.9 pp, low confidence
MKA41.9%estimated ± 5.7 pp, low confidence
SkillsBench official leaderboard (BenchFlow)50.1%estimated ± 6.3 pp, low confidence
CompileBench84.4%estimated ± 10.0 pp, low confidence
OTelBench22.6%estimated ± 4.1 pp, medium confidence
Vibe Code Bench0.2%estimated ± 8.4 pp, medium confidence
FACTS Multimodal39.3%estimated ± 4.7 pp, medium confidence
Hermes Bench62.0%estimated ± 2.9 pp, high confidence
Hermes Index37.2%estimated ± 3.1 pp, high confidence
ExploitBench36.7%estimated ± 7.2 pp, low confidence
ExploitGym8.8%estimated ± 2.7 pp, medium confidence
C-Eval92.3%estimated ± 0.9 pp, medium confidence
Claw-Eval62.7%estimated ± 6.4 pp, low confidence
DeepPlanning35.4%estimated ± 8.3 pp, low confidence
INCLUDE83.7%estimated ± 1.4 pp, low confidence
MAXIFE77.3%estimated ± 6.5 pp, low confidence
MCP Atlas65.7%estimated ± 7.8 pp, low confidence
MCP-Tasks75.1%estimated ± 3.6 pp, low confidence
MMLU-ProX83.0%measured
MMLU-Redux94.1%estimated ± 0.6 pp, high confidence
NL2Repo42.6%estimated ± 3.4 pp, high confidence
NOVA-6356.7%estimated ± 1.3 pp, low confidence
PolyMath49.0%estimated ± 11.4 pp, low confidence
PostTrainBench v110.7%estimated ± 3.6 pp, medium confidence
QwenClawBench86.4%estimated ± 6.0 pp, low confidence
SuperGPQA67.4%estimated ± 7.2 pp, medium confidence
Terminal-Bench 2.059.4%estimated ± 5.8 pp, medium confidence
Toolathlon21.2%estimated ± 4.2 pp, medium confidence
AA LiveCodeBench86.3%estimated ± 2.6 pp, high confidence
HealthBench Hard0.0%estimated ± 5.1 pp, low confidence
MedXpertQA (Text)45.9%estimated ± 4.5 pp, low confidence
MMLU-Pro (Arcee)85.8%estimated ± 3.4 pp, medium confidence
React Native Evals76.0%estimated ± 2.7 pp, medium confidence
ResearchClawBench16.9%estimated ± 1.9 pp, high confidence
SWE-Rebench57.0%estimated ± 3.6 pp, medium confidence
PostTrainBench12.1%estimated ± 3.7 pp, medium confidence
DAYJOB0.0%estimated ± 3.2 pp, medium confidence
DAYJOB Healthcare0.7%estimated ± 4.2 pp, high confidence
Finance Agent v256.1%estimated ± 3.8 pp, medium confidence
BioMysteryBench (human-solvable)81.7%estimated ± 0.6 pp, low confidence
HealthBench (raw)37.2%estimated ± 6.0 pp, low confidence
HealthBench (length-adjusted)49.2%estimated ± 3.4 pp, low confidence
HealthBench Professional36.6%estimated ± 4.2 pp, low confidence
HealthBench Professional (raw)48.5%estimated ± 4.5 pp, medium confidence
HLE-Verified2.3%estimated ± 4.6 pp, low confidence
LABBench280.8%estimated ± 1.8 pp, low confidence
ScarfBench24.6%estimated ± 9.1 pp, low confidence
BlenderWorld (text-only)70.1%estimated ± 2.8 pp, medium confidence
Cookie-Bench77.1%estimated ± 1.3 pp, medium confidence
STEMBench2.3%estimated ± 0.2 pp, low confidence
SimpleQA20.3%estimated ± 4.3 pp, medium confidence
LiveCodeBench74.2%estimated ± 14.6 pp, low confidence
Agents' Last Exam35.4%estimated ± 11.0 pp, low confidence
OpenHarmony Bench v1.051.0%estimated ± 2.4 pp, medium confidence
SEC-Bench Pro17.2%estimated ± 9.3 pp, low confidence
MRCRv258.0%estimated ± 11.3 pp, low confidence
MMLongBench-Doc60.8%estimated ± 2.6 pp, low confidence
BioBasicsBench33.9%estimated ± 2.8 pp, low confidence
LifeVal9.3%estimated ± 1.6 pp, low confidence
ExploitBench Internal Port0.1%estimated ± 5.2 pp, low confidence
PinchBench90.0%measured