Noometry

Model comparison

GPT-6 Luna vs Llama 3-70B

GPT-6 Luna is the stronger model overall, scoring 53.3 to 28.8 on the Noometry Index.

Last verified . 21 shared benchmarks.

GPT-6 Luna OpenAI

53.3

Rank #36 Confirmed

Llama 3-70B Meta

28.8

Rank #323 Confirmed

Summary

  • They share 21 benchmarks with published results for both. GPT-6 Luna scores higher in 9 categories and Llama 3-70B in 0 categories; 9 gaps are clear of the uncertainty.
  • The widest gap is in math, where GPT-6 Luna leads 76.1 to 12.8.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 98.9% for GPT-6 Luna and 4.3% for Llama 3-70B.
  • Llama 3-70B has downloadable open weights; the other is API-only.

Side by side

GPT-6 LunaLlama 3-70B
ProviderOpenAIMeta
Noometry Index53.328.8
Released2026-09-222024-04-18
WeightsProprietaryOpen
Context window1.05M—
Max output128K—
Input $ / M tokens$0.10—
Output $ / M tokens$0.50—
Results tracked4231

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

Category by category

Coding GPT-6 Luna leads

GPT-6 Luna: 55.5 (#25), Llama 3-70B: 35.8 (#218)

BenchmarkGPT-6 LunaLlama 3-70B
LMArena Coding14391206
DeepSWE66.6%—
FrontierCode42.4%—
LMArena WebDev1581—
SciCode54.6%—
BigCodeBench Instruct—43.6%
BigCodeBench Complete—54.5%
ALE-Bench1,577—
HumanEval+—72%
MBPP+—69%

Agentic & Tool Use GPT-6 Luna leads

GPT-6 Luna: 33.3 (#54), Llama 3-70B: 21.1 (#139)

BenchmarkGPT-6 LunaLlama 3-70B
APEX-Agents44.3%—
Cybench—5%
GDP.pdf23%—

Reasoning GPT-6 Luna leads

GPT-6 Luna: 48.2 (#41), Llama 3-70B: 18.0 (#288)

BenchmarkGPT-6 LunaLlama 3-70B
LMArena Hard Prompts14111195
DTBench90.1%54.2%
Epoch Capabilities Index156.28122.93
ARC-AGI-259.3%—
Kagi LLM Benchmark—35.1%
NYT Connections (extended)68.7%—
ARC-AGI-186.7%—
CritPt19.4%—
Chess Puzzles31%—
Mystery Game Puzzles7%—
LMCA44.5%—
ForecastBench—57.1
WinoGrande—83.5%

Math GPT-6 Luna leads

GPT-6 Luna: 76.1 (#15), Llama 3-70B: 12.8 (#305)

BenchmarkGPT-6 LunaLlama 3-70B
OTIS Mock AIME 2024-202598.9%4.3%
LMArena Math14161218
FrontierMath (Tiers 1-3)78.9%—
FrontierMath Tier 456.1%—
ProofBench64%—
MATH Level 5—22.6%

Knowledge GPT-6 Luna leads

GPT-6 Luna: 57.0 (#41), Llama 3-70B: 20.8 (#277)

BenchmarkGPT-6 LunaLlama 3-70B
GPQA Diamond90.5%40.6%
LMArena Expert14441149
SimpleQA Verified41.4%—
MMLU—79.3%

Multimodal Not comparable

GPT-6 Luna: 42.4 (#30), Llama 3-70B: —

BenchmarkGPT-6 LunaLlama 3-70B
LMArena Vision1217—
Blueprint-Bench 231.2%—
Furniture Assembly44.2%—

Multilingual GPT-6 Luna leads

GPT-6 Luna: 50.5 (#117), Llama 3-70B: 33.6 (#251)

BenchmarkGPT-6 LunaLlama 3-70B
LMArena Non-English13861142
LMArena Chinese14331114
LMArena French14201232
LMArena German13691169
LMArena Japanese13691017
LMArena Korean13601017
LMArena Russian13941159
LMArena Spanish13931241

Instruction Following GPT-6 Luna leads

GPT-6 Luna: 74.3 (#99), Llama 3-70B: 62.5 (#238)

BenchmarkGPT-6 LunaLlama 3-70B
LMArena Instruction Following14091194

Long Context GPT-6 Luna leads

GPT-6 Luna: 43.0 (#111), Llama 3-70B: 35.6 (#240)

BenchmarkGPT-6 LunaLlama 3-70B
LMArena Longer Query14091174

Writing & Preference GPT-6 Luna leads

GPT-6 Luna: 58.3 (#119), Llama 3-70B: 42.8 (#231)

BenchmarkGPT-6 LunaLlama 3-70B
LMArena Text13911221
LMArena Creative Writing13631210
LMArena Multi-Turn13961223

Frequently asked questions

Is GPT-6 Luna better than Llama 3-70B?

GPT-6 Luna is the stronger model overall, scoring 53.3 to 28.8 on the Noometry Index.

Is GPT-6 Luna or Llama 3-70B better for coding?

GPT-6 Luna scores higher on coding benchmarks: 55.5 versus 35.8 in the Noometry coding category.

How many benchmarks do GPT-6 Luna and Llama 3-70B share?

21 benchmarks have published results for both models. GPT-6 Luna has 42 scored results on Noometry and Llama 3-70B has 31.

Related comparisons

Go deeper