Noometry

Model comparison

Deepseek Coder v2 vs Gemma 3 27B

Deepseek Coder v2 is the stronger model overall, scoring 35.3 to 30.5 on the Noometry Index.

Last verified . 17 shared benchmarks.

Deepseek Coder v2 DeepSeek

35.3

Rank #217 Confirmed

Gemma 3 27B Google

30.5

Rank #277 Confirmed

Summary

  • They share 17 benchmarks with published results for both. Deepseek Coder v2 scores higher in 5 categories and Gemma 3 27B in 3 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in coding, where Deepseek Coder v2 leads 37.5 to 22.1.

Side by side

Deepseek Coder v2Gemma 3 27B
ProviderDeepSeekGoogle
Noometry Index35.330.5
Released2024-06-172025-03-11
WeightsOpenOpen
Context window—131K
Max output—8K
Input $ / M tokens—$0.08
Output $ / M tokens—$0.16
Results tracked2443

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

Category by category

Coding Deepseek Coder v2 leads

Deepseek Coder v2: 37.5 (#178), Gemma 3 27B: 22.1 (#329)

BenchmarkDeepseek Coder v2Gemma 3 27B
LMArena Coding12511322
Aider Polyglot—4.9%
SciCode—21.2%
BigCodeBench Instruct48.2%—
LiveBench Coding—39.9%
BigCodeBench Complete59.7%—
HumanEval+82.3%—
MBPP+75.1%—

Agentic & Tool Use Not comparable

Deepseek Coder v2: —, Gemma 3 27B: 24.9 (#105)

BenchmarkDeepseek Coder v2Gemma 3 27B
Berkeley Function Calling Leaderboard—29.5%

Reasoning Deepseek Coder v2 leads

Deepseek Coder v2: 22.6 (#175), Gemma 3 27B: 16.3 (#297)

BenchmarkDeepseek Coder v2Gemma 3 27B
LMArena Hard Prompts12071340
Kagi LLM Benchmark—40.4%
CritPt—0%
Chess Puzzles—0%
LiveBench Reasoning—43.8%
DTBench—52.5%
LiveBench Data Analysis—51.5%
LMCA—12.3%
Epoch Capabilities Index—130.04
LiveBench—50%
WinoGrande83.7%—

Math Deepseek Coder v2 leads

Deepseek Coder v2: 32.3 (#188), Gemma 3 27B: 24.4 (#261)

BenchmarkDeepseek Coder v2Gemma 3 27B
LMArena Math12411312
OTIS Mock AIME 2024-2025—22.5%
LiveBench Math—55.4%
MATH Level 5—74%
GSM8K94.5%—

Knowledge Deepseek Coder v2 leads

Deepseek Coder v2: 31.7 (#209), Gemma 3 27B: 25.4 (#256)

BenchmarkDeepseek Coder v2Gemma 3 27B
LMArena Expert11811303
GPQA Diamond—47.7%
Confabulations—40.3%
Vectara Hallucination Rate—7.4%
ARC (AI2) Challenge64.3%—

Multimodal Not comparable

Deepseek Coder v2: —, Gemma 3 27B: 32.0 (#94)

BenchmarkDeepseek Coder v2Gemma 3 27B
LMArena Vision—1165
GeoBench—52%

Multilingual Gemma 3 27B leads

Deepseek Coder v2: 36.4 (#235), Gemma 3 27B: 47.0 (#150)

BenchmarkDeepseek Coder v2Gemma 3 27B
LMArena Non-English11821334
LMArena Chinese12011347
LMArena French11841368
LMArena German11641362
LMArena Japanese11261287
LMArena Korean11041308
LMArena Russian11881349
LMArena Spanish11531348

Instruction Following Gemma 3 27B leads

Deepseek Coder v2: 61.9 (#237), Gemma 3 27B: 70.8 (#155)

BenchmarkDeepseek Coder v2Gemma 3 27B
LMArena Instruction Following11801321
LiveBench Instruction Following—74.9%

Long Context Deepseek Coder v2 leads

Deepseek Coder v2: 37.0 (#219), Gemma 3 27B: 27.6 (#288)

BenchmarkDeepseek Coder v2Gemma 3 27B
LMArena Longer Query12191333
Fiction.LiveBench—33.3%

Writing & Preference Gemma 3 27B leads

Deepseek Coder v2: 38.4 (#249), Gemma 3 27B: 52.7 (#161)

BenchmarkDeepseek Coder v2Gemma 3 27B
LMArena Text11911358
LMArena Creative Writing11201346
LMArena Multi-Turn11761344
Short-Story Creative Writing—79.9%
EQ-Bench Creative Writing—1266
LiveBench Language—34.6%

Frequently asked questions

Is Deepseek Coder v2 better than Gemma 3 27B?

Deepseek Coder v2 is the stronger model overall, scoring 35.3 to 30.5 on the Noometry Index.

Is Deepseek Coder v2 or Gemma 3 27B better for coding?

Deepseek Coder v2 scores higher on coding benchmarks: 37.5 versus 22.1 in the Noometry coding category.

How many benchmarks do Deepseek Coder v2 and Gemma 3 27B share?

17 benchmarks have published results for both models. Deepseek Coder v2 has 24 scored results on Noometry and Gemma 3 27B has 43.

Related comparisons

Go deeper