Noometry

Model comparison

GLM-5.3-Flash vs Llama 2-70B

GLM-5.3-Flash is the stronger model overall, scoring 51.8 to 24.4 on the Noometry Index.

Last verified . 20 shared benchmarks.

GLM-5.3-Flash Z.ai (Zhipu)

51.8

Rank #41 Confirmed

Llama 2-70B Meta

24.4

Rank #349 Confirmed

Summary

  • They share 20 benchmarks with published results for both. GLM-5.3-Flash scores higher in 8 categories and Llama 2-70B in 0 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in knowledge, where GLM-5.3-Flash leads 58.4 to 7.4.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 93.9% for GLM-5.3-Flash and 0% for Llama 2-70B.

Side by side

GLM-5.3-FlashLlama 2-70B
ProviderZ.ai (Zhipu)Meta
Noometry Index51.824.4
Released2026-08-202023-07-18
WeightsOpenOpen
Context window1M—
Max output131K—
Input $ / M tokens$0.15—
Output $ / M tokens$0.50—
Results tracked4035

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

Category by category

Coding GLM-5.3-Flash leads

GLM-5.3-Flash: 53.1 (#31), Llama 2-70B: 31.4 (#286)

BenchmarkGLM-5.3-FlashLlama 2-70B
LMArena Coding15081079
DeepSWE63.4%—
FrontierCode31.8%—
CursorBench36.8%—
LMArena WebDev1609—
FrontierSWE18.1%—
SciCode51.6%—
ALE-Bench303.55—

Agentic & Tool Use Not comparable

GLM-5.3-Flash: 34.2 (#47), Llama 2-70B: —

BenchmarkGLM-5.3-FlashLlama 2-70B
APEX-Agents52.8%—
GDP.pdf14%—

Reasoning GLM-5.3-Flash leads

GLM-5.3-Flash: 48.0 (#42), Llama 2-70B: 14.4 (#325)

BenchmarkGLM-5.3-FlashLlama 2-70B
LMArena Hard Prompts14911073
Epoch Capabilities Index151.88113.79
ARC-AGI-265.8%—
ARC-AGI-191%—
CritPt15.4%—
Chess Puzzles14%—
Mystery Game Puzzles8%—
DTBench—41.6%
Surface Evolver Bench52.5%—
Bench to the Future 30.15—
BIG-Bench Hard—64.9%
CommonsenseQA 2.0—50%
ForecastBench—51.4
HellaSwag—85.3%
LAMBADA—78.9%
PIQA—82.8%
WinoGrande—80.2%

Math GLM-5.3-Flash leads

GLM-5.3-Flash: 53.3 (#47), Llama 2-70B: 8.1 (#326)

BenchmarkGLM-5.3-FlashLlama 2-70B
OTIS Mock AIME 2024-202593.9%0%
LMArena Math15001091
FrontierMath (Tiers 1-3)55.8%—
FrontierMath Tier 417.1%—
ProofBench21%—
MATH Level 5—3.3%
GSM8K—69.6%

Knowledge GLM-5.3-Flash leads

GLM-5.3-Flash: 58.4 (#36), Llama 2-70B: 7.4 (#310)

BenchmarkGLM-5.3-FlashLlama 2-70B
GPQA Diamond90.2%26.3%
LMArena Expert15131039
ARC (AI2) Challenge—78.3%
BoolQ—88.6%
MMLU—69.9%
OpenBookQA—60.2%
TriviaQA—87.6%

Multimodal Not comparable

GLM-5.3-Flash: 42.8 (#27), Llama 2-70B: —

BenchmarkGLM-5.3-FlashLlama 2-70B
LMArena Vision1296—

Multilingual GLM-5.3-Flash leads

GLM-5.3-Flash: 56.0 (#25), Llama 2-70B: 27.7 (#274)

BenchmarkGLM-5.3-FlashLlama 2-70B
LMArena Non-English14621045
LMArena Chinese1527995
LMArena French14961090
LMArena German14701041
LMArena Japanese1429927
LMArena Korean1446964
LMArena Russian14691083
LMArena Spanish14711143

Instruction Following GLM-5.3-Flash leads

GLM-5.3-Flash: 77.5 (#20), Llama 2-70B: 54.9 (#278)

BenchmarkGLM-5.3-FlashLlama 2-70B
LMArena Instruction Following14781071

Long Context GLM-5.3-Flash leads

GLM-5.3-Flash: 45.4 (#39), Llama 2-70B: 32.3 (#270)

BenchmarkGLM-5.3-FlashLlama 2-70B
LMArena Longer Query14821062

Writing & Preference GLM-5.3-Flash leads

GLM-5.3-Flash: 65.3 (#50), Llama 2-70B: 32.3 (#279)

BenchmarkGLM-5.3-FlashLlama 2-70B
LMArena Text14711115
LMArena Creative Writing14421075
LMArena Multi-Turn14671088

Frequently asked questions

Is GLM-5.3-Flash better than Llama 2-70B?

GLM-5.3-Flash is the stronger model overall, scoring 51.8 to 24.4 on the Noometry Index.

Is GLM-5.3-Flash or Llama 2-70B better for coding?

GLM-5.3-Flash scores higher on coding benchmarks: 53.1 versus 31.4 in the Noometry coding category.

How many benchmarks do GLM-5.3-Flash and Llama 2-70B share?

20 benchmarks have published results for both models. GLM-5.3-Flash has 40 scored results on Noometry and Llama 2-70B has 35.

Related comparisons

Go deeper