Noometry

Model comparison

GPT-4 Turbo vs Qwen2-72B

GPT-4 Turbo and Qwen2-72B score almost the same on the Noometry Index (30.5 vs 30.0), so choose on price, context window or the category you care about most.

Last verified . 25 shared benchmarks.

GPT-4 Turbo OpenAI

30.5

Rank #292 Confirmed

Qwen2-72B Alibaba (Qwen)

30.0

Rank #300 Confirmed

Summary

  • They share 25 benchmarks with published results for both. GPT-4 Turbo scores higher in 6 categories and Qwen2-72B in 2 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in math, where Qwen2-72B leads 30.2 to 9.0.
  • The biggest single-benchmark swing is BigCodeBench Instruct: 48.2% for GPT-4 Turbo and 38.5% for Qwen2-72B.
  • Qwen2-72B has downloadable open weights; the other is API-only.

Side by side

GPT-4 TurboQwen2-72B
ProviderOpenAIAlibaba (Qwen)
Noometry Index30.530.0
Released2023-11-062024-06-07
WeightsProprietaryOpen
Context window128K—
Max output4K—
Input $ / M tokens$10—
Output $ / M tokens$30—
Results tracked3626

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

Category by category

Coding GPT-4 Turbo leads

GPT-4 Turbo: 33.8 (#249), Qwen2-72B: 29.1 (#310)

BenchmarkGPT-4 TurboQwen2-72B
WeirdML18%11.3%
BigCodeBench Instruct48.2%38.5%
LMArena Coding12681196
BigCodeBench Complete58.2%54%
HumanEval+86.6%—
MBPP+73.3%—

Agentic & Tool Use Not comparable

GPT-4 Turbo: —, Qwen2-72B: 17.0 (#146)

BenchmarkGPT-4 TurboQwen2-72B
METR Time Horizons36.7%29.9%
TheAgentCompany—1.1%

Reasoning Qwen2-72B leads

GPT-4 Turbo: 15.3 (#317), Qwen2-72B: 23.2 (#181)

BenchmarkGPT-4 TurboQwen2-72B
LMArena Hard Prompts12511191
Epoch Capabilities Index127.25125.28
SimpleBench25.1%—
Chess Puzzles6%—
DTBench61.6%—
LMCA9.8%—
ForecastBench59.4—

Math Qwen2-72B leads

GPT-4 Turbo: 9.0 (#322), Qwen2-72B: 30.2 (#236)

BenchmarkGPT-4 TurboQwen2-72B
LMArena Math12721235
MATH Level 546.7%39.1%
FrontierMath (Tiers 1-3)0.7%—
OTIS Mock AIME 2024-20256.7%—

Knowledge GPT-4 Turbo leads

GPT-4 Turbo: 24.3 (#268), Qwen2-72B: 21.2 (#275)

BenchmarkGPT-4 TurboQwen2-72B
GPQA Diamond46.6%40.8%
LMArena Expert12231171
MMLU81.3%82.4%
Confabulations28.4%—

Multimodal Not comparable

GPT-4 Turbo: 30.6 (#110), Qwen2-72B: —

BenchmarkGPT-4 TurboQwen2-72B
LMArena Vision1090—

Multilingual GPT-4 Turbo leads

GPT-4 Turbo: 40.5 (#216), Qwen2-72B: 35.9 (#244)

BenchmarkGPT-4 TurboQwen2-72B
LMArena Non-English12451176
LMArena Chinese12421240
LMArena French12761170
LMArena German12591151
LMArena Japanese11941111
LMArena Korean11871083
LMArena Russian12591169
LMArena Spanish12601169

Instruction Following GPT-4 Turbo leads

GPT-4 Turbo: 65.8 (#216), Qwen2-72B: 61.7 (#241)

BenchmarkGPT-4 TurboQwen2-72B
LMArena Instruction Following12491181

Long Context GPT-4 Turbo leads

GPT-4 Turbo: 38.0 (#206), Qwen2-72B: 36.1 (#235)

BenchmarkGPT-4 TurboQwen2-72B
LMArena Longer Query12541192

Writing & Preference GPT-4 Turbo leads

GPT-4 Turbo: 47.7 (#206), Qwen2-72B: 40.8 (#241)

BenchmarkGPT-4 TurboQwen2-72B
LMArena Text12721203
LMArena Creative Writing12691181
LMArena Multi-Turn12671196

Frequently asked questions

Is GPT-4 Turbo better than Qwen2-72B?

GPT-4 Turbo and Qwen2-72B score almost the same on the Noometry Index (30.5 vs 30.0), so choose on price, context window or the category you care about most.

Is GPT-4 Turbo or Qwen2-72B better for coding?

GPT-4 Turbo scores higher on coding benchmarks: 33.8 versus 29.1 in the Noometry coding category.

How many benchmarks do GPT-4 Turbo and Qwen2-72B share?

25 benchmarks have published results for both models. GPT-4 Turbo has 36 scored results on Noometry and Qwen2-72B has 26.

Related comparisons

Go deeper