Noometry

Model comparison

Grok 3 vs Nemotron 3 Ultra

Nemotron 3 Ultra is the stronger model overall, scoring 42.5 to 39.9 on the Noometry Index.

Last verified . 21 shared benchmarks.

Grok 3 xAI

39.9

Rank #157 Confirmed

Nemotron 3 Ultra NVIDIA

42.5

Rank #113 Confirmed

Summary

  • They share 21 benchmarks with published results for both. Grok 3 scores higher in 4 categories and Nemotron 3 Ultra in 5 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in reasoning, where Nemotron 3 Ultra leads 31.0 to 13.7.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 55.6% for Grok 3 and 86.7% for Nemotron 3 Ultra.
  • Nemotron 3 Ultra has downloadable open weights; the other is API-only.

Side by side

Grok 3Nemotron 3 Ultra
ProviderxAINVIDIA
Noometry Index39.942.5
Released2025-04-092026-06-04
WeightsProprietaryOpen
Context window—262K
Max output—128K
Input $ / M tokens—$0.50
Output $ / M tokens—$2.20
Results tracked4030

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding Grok 3 leads

Grok 3: 41.9 (#115), Nemotron 3 Ultra: 38.1 (#182)

BenchmarkGrok 3Nemotron 3 Ultra
WeirdML37.2%43.5%
LMArena Coding14321468
FrontierCode—13.6%
Aider Polyglot53.3%—
SciCode—40.3%

Agentic & Tool Use Grok 3 leads

Grok 3: 30.5 (#76), Nemotron 3 Ultra: 23.4 (#126)

BenchmarkGrok 3Nemotron 3 Ultra
APEX-Agents—22.7%
BALROG29.5%—

Reasoning Nemotron 3 Ultra leads

Grok 3: 13.7 (#333), Nemotron 3 Ultra: 31.0 (#82)

BenchmarkGrok 3Nemotron 3 Ultra
LMArena Hard Prompts14341452
Epoch Capabilities Index138.33146.17
ARC-AGI-20%—
SimpleBench36.1%—
Kagi LLM Benchmark61.3%—
ARC-AGI-15.5%—
CritPt—3.1%
Chess Puzzles—12%
Mystery Game Puzzles—20%
DTBench—90.1%
LMCA—36.9%

Math Grok 3 leads

Grok 3: 38.0 (#145), Nemotron 3 Ultra: 35.3 (#187)

BenchmarkGrok 3Nemotron 3 Ultra
OTIS Mock AIME 2024-202555.6%86.7%
LMArena Math13911457
ProofBench—2%
Omni-MATH46.4%—
MATH Level 588.7%—
FrontierMath (Feb 2025 set)3.8%—
FrontierMath Tier 4 (v1)0%—

Knowledge Nemotron 3 Ultra leads

Grok 3: 46.2 (#82), Nemotron 3 Ultra: 52.5 (#61)

BenchmarkGrok 3Nemotron 3 Ultra
GPQA Diamond75.8%85.4%
LMArena Expert14211472
MMLU-Pro78.8%—
Confabulations14.2%—
Vectara Hallucination Rate5.8%—
GPQA (HELM)65%—

Multilingual Nemotron 3 Ultra leads

Grok 3: 52.3 (#87), Nemotron 3 Ultra: 53.5 (#64)

BenchmarkGrok 3Nemotron 3 Ultra
LMArena Non-English14101427
LMArena Chinese14481497
LMArena French14601472
LMArena German14311471
LMArena Korean13731386
LMArena Russian14161417
LMArena Spanish14171454
LMArena Japanese1387—

Instruction Following Too close to call

Grok 3: 75.0 (#73), Nemotron 3 Ultra: 74.7 (#88)

BenchmarkGrok 3Nemotron 3 Ultra
LMArena Instruction Following14091418
IFEval88.4%—

Long Context Nemotron 3 Ultra leads

Grok 3: 38.7 (#192), Nemotron 3 Ultra: 43.9 (#81)

BenchmarkGrok 3Nemotron 3 Ultra
LMArena Longer Query14391437
Fiction.LiveBench58.3%—

Writing & Preference Nemotron 3 Ultra leads

Grok 3: 55.8 (#141), Nemotron 3 Ultra: 66.3 (#36)

BenchmarkGrok 3Nemotron 3 Ultra
LMArena Text14261445
LMArena Creative Writing14141398
EQ-Bench Creative Writing11861692
LMArena Multi-Turn14251411
Short-Story Creative Writing76.4%—
WildBench84.9%—

Frequently asked questions

Is Grok 3 better than Nemotron 3 Ultra?

Nemotron 3 Ultra is the stronger model overall, scoring 42.5 to 39.9 on the Noometry Index.

Is Grok 3 or Nemotron 3 Ultra better for coding?

Grok 3 scores higher on coding benchmarks: 41.9 versus 38.1 in the Noometry coding category.

How many benchmarks do Grok 3 and Nemotron 3 Ultra share?

21 benchmarks have published results for both models. Grok 3 has 40 scored results on Noometry and Nemotron 3 Ultra has 30.

Related comparisons

Go deeper