Noometry

Model comparison

Gemma 2 27B vs GPT-4

Gemma 2 27B and GPT-4 score almost the same on the Noometry Index (29.4 vs 29.1), so choose on price, context window or the category you care about most.

Last verified . 26 shared benchmarks.

Gemma 2 27B Google

29.4

Rank #312 Confirmed

GPT-4 OpenAI

29.1

Rank #316 Confirmed

Summary

  • They share 26 benchmarks with published results for both. Gemma 2 27B scores higher in 3 categories and GPT-4 in 5 categories; 5 gaps are clear of the uncertainty.
  • The widest gap is in writing & preference, where Gemma 2 27B leads 44.2 to 34.9.
  • The biggest single-benchmark swing is DTBench: 48% for Gemma 2 27B and 62.7% for GPT-4.
  • Gemma 2 27B is cheaper at $0.65 / $0.65 per million input/output tokens, against $30 / $60 for GPT-4.
  • Gemma 2 27B has downloadable open weights; the other is API-only.

Side by side

Gemma 2 27BGPT-4
ProviderGoogleOpenAI
Noometry Index29.429.1
Released2024-06-242023-03-14
WeightsOpenProprietary
Context window8K8K
Max output2K8K
Input $ / M tokens$0.65$30
Output $ / M tokens$0.65$60
Results tracked3438

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

Category by category

Coding Gemma 2 27B leads

Gemma 2 27B: 34.1 (#246), GPT-4: 31.6 (#283)

BenchmarkGemma 2 27BGPT-4
BigCodeBench Instruct42.8%46%
LMArena Coding12111254
BigCodeBench Complete52.5%57.2%
WeirdML—12.4%
LiveBench Coding36%—
HumanEval+—79.3%

Agentic & Tool Use Not comparable

Gemma 2 27B: —, GPT-4: —

BenchmarkGemma 2 27BGPT-4
METR Time Horizons—36.1%

Reasoning GPT-4 leads

Gemma 2 27B: 15.3 (#315), GPT-4: 17.8 (#289)

BenchmarkGemma 2 27BGPT-4
LMArena Hard Prompts11981241
DTBench48%62.7%
LMCA7.1%17.1%
Epoch Capabilities Index122.08125.89
Chess Puzzles—4%
LiveBench Reasoning28.1%—
Mystery Game Puzzles—12%
LiveBench Data Analysis47.9%—
BIG-Bench Hard—75.1%
ForecastBench—57.8
HellaSwag—95.3%
LiveBench38.2%—
WinoGrande—87.5%

Math Too close to call

Gemma 2 27B: 10.7 (#311), GPT-4: 10.8 (#309)

BenchmarkGemma 2 27BGPT-4
OTIS Mock AIME 2024-20251.4%1.1%
LMArena Math12121269
MATH Level 527.9%23%
LiveBench Math26.5%—
GSM8K—92%

Knowledge Too close to call

Gemma 2 27B: 19.0 (#280), GPT-4: 18.4 (#282)

BenchmarkGemma 2 27BGPT-4
GPQA Diamond36.5%35.7%
LMArena Expert11721211
MMLU75.7%86.4%
Confabulations27.1%—
TriviaQA—84.8%

Multilingual GPT-4 leads

Gemma 2 27B: 38.6 (#226), GPT-4: 40.6 (#215)

BenchmarkGemma 2 27BGPT-4
LMArena Non-English12171246
LMArena Chinese12211242
LMArena French12471283
LMArena German12091251
LMArena Japanese11751209
LMArena Korean11741184
LMArena Russian12341251
LMArena Spanish12281261

Instruction Following GPT-4 leads

Gemma 2 27B: 60.5 (#249), GPT-4: 65.3 (#222)

BenchmarkGemma 2 27BGPT-4
LMArena Instruction Following12061241
LiveBench Instruction Following58.1%—

Long Context Too close to call

Gemma 2 27B: 37.3 (#218), GPT-4: 37.7 (#212)

BenchmarkGemma 2 27BGPT-4
LMArena Longer Query12311244

Writing & Preference Gemma 2 27B leads

Gemma 2 27B: 44.2 (#225), GPT-4: 34.9 (#268)

BenchmarkGemma 2 27BGPT-4
LMArena Text12311263
LMArena Creative Writing12411244
LMArena Multi-Turn12241257
EQ-Bench Creative Writing—752
LiveBench Language32.6%—

Frequently asked questions

Is Gemma 2 27B better than GPT-4?

Gemma 2 27B and GPT-4 score almost the same on the Noometry Index (29.4 vs 29.1), so choose on price, context window or the category you care about most.

Which is cheaper, Gemma 2 27B or GPT-4?

Gemma 2 27B is cheaper. It lists at $0.65 per million input tokens and $0.65 per million output tokens; GPT-4 lists at $30 and $60.

Is Gemma 2 27B or GPT-4 better for coding?

Gemma 2 27B scores higher on coding benchmarks: 34.1 versus 31.6 in the Noometry coding category.

Which has the bigger context window?

Both accept 8K tokens.

How many benchmarks do Gemma 2 27B and GPT-4 share?

26 benchmarks have published results for both models. Gemma 2 27B has 34 scored results on Noometry and GPT-4 has 38.

Related comparisons

Go deeper