Noometry

Model comparison

Gemma 2 27B vs Llama 3-8B

Gemma 2 27B is the stronger model overall, scoring 29.2 to 25.3 on the Noometry Index. Llama 3-8B costs 4.6× less per token, which makes it the better buy when Gemma 2 27B's lead doesn't matter for your workload.

Last verified . 25 shared benchmarks.

Gemma 2 27B Google

29.2

Rank #302 Confirmed

Llama 3-8B Meta

25.3

Rank #338 Confirmed

Summary

  • They share 25 benchmarks with published results for both. Gemma 2 27B scores higher in 8 categories and Llama 3-8B in 0 categories; 7 gaps are clear of the uncertainty.
  • The widest gap is in knowledge, where Gemma 2 27B leads 18.9 to 7.8.
  • The biggest single-benchmark swing is MATH Level 5: 27.9% for Gemma 2 27B and 6.1% for Llama 3-8B.
  • Llama 3-8B is cheaper at $0.14 / $0.14 per million input/output tokens, against $0.65 / $0.65 for Gemma 2 27B.

Side by side

Gemma 2 27BLlama 3-8B
ProviderGoogleMeta
Noometry Index29.225.3
Released2024-06-242024-04-18
WeightsOpenOpen
Context window8K8K
Max output2K8K
Input $ / M tokens$0.65$0.14
Output $ / M tokens$0.65$0.14
Results tracked3434

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

Category by category

Coding Gemma 2 27B leads

Gemma 2 27B: 33.5 (#243), Llama 3-8B: 30.5 (#285)

BenchmarkGemma 2 27BLlama 3-8B
BigCodeBench Instruct42.8%31.9%
LMArena Coding12111152
BigCodeBench Complete52.5%36.9%
LiveBench Coding36%—
HumanEval+—56.7%
MBPP+—54.8%

Reasoning Too close to call

Gemma 2 27B: 14.8 (#313), Llama 3-8B: 13.8 (#321)

BenchmarkGemma 2 27BLlama 3-8B
LMArena Hard Prompts11991133
DTBench48%43.9%
Epoch Capabilities Index122.06116.33
Chess Puzzles—0%
LiveBench Reasoning28.1%—
LiveBench Data Analysis47.9%—
LMCA7.1%—
Adversarial NLI—57.3%
ForecastBench—58.6
LiveBench38.2%—
WinoGrande—75.7%

Math Gemma 2 27B leads

Gemma 2 27B: 10.0 (#306), Llama 3-8B: 8.3 (#317)

BenchmarkGemma 2 27BLlama 3-8B
OTIS Mock AIME 2024-20251.4%1.9%
LMArena Math12121151
MATH Level 527.9%6.1%
LiveBench Math26.5%—

Knowledge Gemma 2 27B leads

Gemma 2 27B: 18.9 (#275), Llama 3-8B: 7.8 (#303)

BenchmarkGemma 2 27BLlama 3-8B
GPQA Diamond36.5%26.1%
LMArena Expert11721113
MMLU75.7%68.8%
Confabulations27.1%—
ARC (AI2) Challenge—82.8%
OpenBookQA—82.6%
TriviaQA—67.7%

Multilingual Gemma 2 27B leads

Gemma 2 27B: 38.8 (#221), Llama 3-8B: 31.0 (#256)

BenchmarkGemma 2 27BLlama 3-8B
LMArena Non-English12171098
LMArena Chinese12211076
LMArena French12461158
LMArena German12091104
LMArena Japanese1175967
LMArena Korean11741003
LMArena Russian12341109
LMArena Spanish12281172

Instruction Following Gemma 2 27B leads

Gemma 2 27B: 60.7 (#244), Llama 3-8B: 58.7 (#255)

BenchmarkGemma 2 27BLlama 3-8B
LMArena Instruction Following12061127
LiveBench Instruction Following58.1%—

Long Context Gemma 2 27B leads

Gemma 2 27B: 37.3 (#213), Llama 3-8B: 34.2 (#246)

BenchmarkGemma 2 27BLlama 3-8B
LMArena Longer Query12311128

Writing & Preference Gemma 2 27B leads

Gemma 2 27B: 44.4 (#220), Llama 3-8B: 37.7 (#251)

BenchmarkGemma 2 27BLlama 3-8B
LMArena Text12311166
LMArena Creative Writing12411150
LMArena Multi-Turn12241152
LiveBench Language32.6%—

Frequently asked questions

Is Gemma 2 27B better than Llama 3-8B?

Gemma 2 27B is the stronger model overall, scoring 29.2 to 25.3 on the Noometry Index. Llama 3-8B costs 4.6× less per token, which makes it the better buy when Gemma 2 27B's lead doesn't matter for your workload.

Which is cheaper, Gemma 2 27B or Llama 3-8B?

Llama 3-8B is cheaper. It lists at $0.14 per million input tokens and $0.14 per million output tokens; Gemma 2 27B lists at $0.65 and $0.65.

Is Gemma 2 27B or Llama 3-8B better for coding?

Gemma 2 27B scores higher on coding benchmarks: 33.5 versus 30.5 in the Noometry coding category.

Which has the bigger context window?

Both accept 8K tokens.

How many benchmarks do Gemma 2 27B and Llama 3-8B share?

25 benchmarks have published results for both models. Gemma 2 27B has 34 scored results on Noometry and Llama 3-8B has 34.

Related comparisons

Go deeper