LLMBenchmark.ai

Enter opens the highlighted result. Esc closes.

DeepSeek V3 0324 vs Gemma 3 27B

Updated October 21, 2026. 21 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.

Replace DeepSeek V3 0324 Replace Gemma 3 27B
DeepSeek V3 0324DeepSeek
Coding
42.0
Overall
9.7
Output $/M
$1.14
20
Gemma 3 27BGoogle
Coding
4.9
Overall
4.9
Output $/M
$0.35
1
vs boards won of 21 shared

Decision reading

Pick DeepSeek V3 0324 for coding: it scores 42.0 on the Vibe Coding Index against 4.9 for Gemma 3 27B.

Pick Gemma 3 27B when output tokens cost 3.3x less, or when the benchmark that matters is Creative writing (Lech Mazur), where it leads.

Model ADeepSeek V3 0324DeepSeek42.0/1001 independent benchmarkNo cost per solved taskModel BGemma 3 27BGoogle4.9/1001 independent benchmarkNo cost per solved task
Shared results
21
DeepSeek V3 0324 only
14
Gemma 3 27B only
41
Like-for-like categories
7 of 9

Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.

Which one for your work

Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.

At a glance

Coding index
DeepSeek V3 0324 42.0 from 1 benchmark, Gemma 3 27B 4.9 from 1 benchmark.
Widest category
Agentic work, 18.6 points apart on average over 2 shared benchmarks.
Widest single gap
τ²-Bench Telecom: DeepSeek V3 0324 47.1% against Gemma 3 27B 10.5%, 36.5 points apart on a 0 to 100 scale.
Output price
DeepSeek V3 0324 $1.14 against Gemma 3 27B $0.35 per million output tokens, 3.3x apart.
Context window
DeepSeek V3 0324 takes 164K tokens, Gemma 3 27B 131K.

Benchmark by benchmark

First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.

Boards won, where both have a result
CategoryDeepSeek V3 0324BarsGemma 3 27BTies
Coding40
Intelligence10
Agentic work20
Reasoning30
Math40
Knowledge10
General51
All shared boards201
CodingLike-for-likeFour of these boards feed the Vibe Coding Index40
Coding benchmarks for DeepSeek V3 0324 and Gemma 3 27B
BenchmarkDeepSeek V3 0324Gemma 3 27BLeader
Aider polyglotNot measured4.9%Only Gemma 3 27B measured
SWE-bench Verified42.0%Not measuredOnly DeepSeek V3 0324 measured
AA Coding Index21.210.1DeepSeek V3 0324
LiveCodeBench (Vals AI)65.5%Not measuredOnly DeepSeek V3 0324 measured
LiveCodeBench (Artificial Analysis run)40.5%13.7%DeepSeek V3 0324
SciCode39.0%23.3%DeepSeek V3 0324
Terminal-Bench Hard (Artificial Analysis run)15.2%3.8%DeepSeek V3 0324
WeirdML36.1%Not measuredOnly DeepSeek V3 0324 measured
LMArena Text: codingNot measured1322Only Gemma 3 27B measured
LMArena Text: software and ITNot measured1354Only Gemma 3 27B measured
Neither model has a result on the other 28 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index10
Intelligence benchmarks for DeepSeek V3 0324 and Gemma 3 27B
BenchmarkDeepSeek V3 0324Gemma 3 27BLeader
AA Intelligence Index9.74.9DeepSeek V3 0324
Epoch Capabilities IndexNot measured130.0Only Gemma 3 27B measured
Agentic workLike-for-likePublic boards, shown as published, kept out of the index20
Agentic work benchmarks for DeepSeek V3 0324 and Gemma 3 27B
BenchmarkDeepSeek V3 0324Gemma 3 27BLeader
AA Agentic Index0.80.1DeepSeek V3 0324
τ²-Bench Telecom47.1%10.5%DeepSeek V3 0324
METR time horizons49.6%Not measuredOnly DeepSeek V3 0324 measured
Neither model has a result on the other 36 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index30
Reasoning benchmarks for DeepSeek V3 0324 and Gemma 3 27B
BenchmarkDeepSeek V3 0324Gemma 3 27BLeader
GPQA Diamond61.6%Not measuredOnly DeepSeek V3 0324 measured
GPQA Diamond (Artificial Analysis run)65.5%42.8%DeepSeek V3 0324
Humanity's Last Exam (Artificial Analysis run)4.7%4.4%DeepSeek V3 0324
GPQA Diamond (Epoch AI run)67.6%47.7%DeepSeek V3 0324
SimpleBench27.2%Not measuredOnly DeepSeek V3 0324 measured
LMArena Text: hard promptsNot measured1340Only Gemma 3 27B measured
LMArena Text: expertNot measured1304Only Gemma 3 27B measured
LMArena Text: hard prompts in EnglishNot measured1353Only Gemma 3 27B measured
LMArena Vision: diagramsNot measured1156Only Gemma 3 27B measured
LMArena Vision: homeworkNot measured1160Only Gemma 3 27B measured
Neither model has a result on the other 11 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index40
Math benchmarks for DeepSeek V3 0324 and Gemma 3 27B
BenchmarkDeepSeek V3 0324Gemma 3 27BLeader
AIME52.2%Not measuredOnly DeepSeek V3 0324 measured
MATH 500 (Vals AI)88.6%Not measuredOnly DeepSeek V3 0324 measured
MGSM (Vals AI)91.7%Not measuredOnly DeepSeek V3 0324 measured
AA Math Index41.020.7DeepSeek V3 0324
AIME 2025 (Artificial Analysis run)41.0%20.7%DeepSeek V3 0324
OTIS Mock AIME37.8%22.5%DeepSeek V3 0324
MATH level 575.5%74.0%DeepSeek V3 0324
LMArena Text: mathNot measured1312Only Gemma 3 27B measured
LMArena Text: mathematical workNot measured1327Only Gemma 3 27B measured
Neither model has a result on the other 4 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index10
Knowledge benchmarks for DeepSeek V3 0324 and Gemma 3 27B
BenchmarkDeepSeek V3 0324Gemma 3 27BLeader
MMLU Pro79.5%Not measuredOnly DeepSeek V3 0324 measured
LegalBench (Vals AI)77.7%Not measuredOnly DeepSeek V3 0324 measured
CorpFin (Vals AI)54.7%Not measuredOnly DeepSeek V3 0324 measured
TaxEval (Vals AI)71.1%Not measuredOnly DeepSeek V3 0324 measured
MedQA (Vals AI)82.0%Not measuredOnly DeepSeek V3 0324 measured
MMLU Pro (Artificial Analysis run)81.9%66.9%DeepSeek V3 0324
GeoBenchNot measured52.0%Only Gemma 3 27B measured
LMArena Text: scienceNot measured1383Only Gemma 3 27B measured
LMArena Text: medicine and healthcareNot measured1391Only Gemma 3 27B measured
LMArena Text: legal and governmentNot measured1375Only Gemma 3 27B measured
LMArena Text: business and financeNot measured1360Only Gemma 3 27B measured
Neither model has a result on the other 6 boards here
Web designNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Web design benchmarks for DeepSeek V3 0324 and Gemma 3 27B
BenchmarkDeepSeek V3 0324Gemma 3 27BLeader
Neither model has a result on 9 boards here
GeneralLike-for-likePublic boards, shown as published, kept out of the index51
General benchmarks for DeepSeek V3 0324 and Gemma 3 27B
BenchmarkDeepSeek V3 0324Gemma 3 27BLeader
IFBench41.0%31.8%DeepSeek V3 0324
Long context reasoning (AA-LCR)40.7%7.3%DeepSeek V3 0324
DTBench64.8%52.5%DeepSeek V3 0324
LMCA15.5%12.3%DeepSeek V3 0324
Fiction.LiveBench50.0%33.3%DeepSeek V3 0324
Creative writing (Lech Mazur)77.0%79.9%Gemma 3 27B
LMArena TextNot measured1358Only Gemma 3 27B measured
LMArena Text: instruction followingNot measured1321Only Gemma 3 27B measured
LMArena Text: creative writingNot measured1345Only Gemma 3 27B measured
LMArena VisionNot measured1165Only Gemma 3 27B measured
LMArena Text: EnglishNot measured1377Only Gemma 3 27B measured
LMArena Text: non-EnglishNot measured1334Only Gemma 3 27B measured
LMArena Text: ChineseNot measured1346Only Gemma 3 27B measured
LMArena Text: GermanNot measured1362Only Gemma 3 27B measured
LMArena Text: JapaneseNot measured1287Only Gemma 3 27B measured
LMArena Text: KoreanNot measured1308Only Gemma 3 27B measured
LMArena Text: FrenchNot measured1368Only Gemma 3 27B measured
LMArena Text: SpanishNot measured1349Only Gemma 3 27B measured
LMArena Text: RussianNot measured1349Only Gemma 3 27B measured
LMArena Text: PolishNot measured1345Only Gemma 3 27B measured
LMArena Text: multi-turnNot measured1345Only Gemma 3 27B measured
LMArena Text: longer queriesNot measured1333Only Gemma 3 27B measured
LMArena Text: writing and languageNot measured1347Only Gemma 3 27B measured
LMArena Text: entertainment and mediaNot measured1320Only Gemma 3 27B measured
LMArena Vision: EnglishNot measured1188Only Gemma 3 27B measured
LMArena Vision: ChineseNot measured1164Only Gemma 3 27B measured
LMArena Vision: reading textNot measured1156Only Gemma 3 27B measured
LMArena Vision: humorNot measured1166Only Gemma 3 27B measured
LMArena Vision: captioningNot measured1151Only Gemma 3 27B measured
LMArena Vision: recognizing thingsNot measured1196Only Gemma 3 27B measured
LMArena Vision: creative writingNot measured1177Only Gemma 3 27B measured
Neither model has a result on the other 15 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for DeepSeek V3 0324 and Gemma 3 27B
BenchmarkDeepSeek V3 0324Gemma 3 27BLeader
Neither model has a result on 5 boards here

What each workload costs

Three fixed token mixes at each model's listed price, so per-token rates become a bill.

Chat turn

1K input + 500 output tokens

DeepSeek V3 0324$0.0009Fits in one request
Gemma 3 27B$0.0003Fits in one request

Gemma 3 27B costs less

Repository review

50K input + 3K output tokens

DeepSeek V3 0324$0.018Fits in one request
Gemma 3 27B$0.0050Fits in one request

Gemma 3 27B costs less

Agent loop

200K input + 10K output tokens

DeepSeek V3 0324$0.069Exceeds the context window
Gemma 3 27B$0.019Exceeds the context window

Gemma 3 27B costs less

Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.

Specification differences

The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.

Specifications of DeepSeek V3 0324 and Gemma 3 27B
SpecificationDeepSeek V3 0324Google Gemma 3 27B
MakerDeepSeekGoogle
WeightsOpen weightsOpen weights
ReleasedMarch 24, 2025March 12, 2025
Context window164K131K
Input $/Mtok$0.29$0.08
Output $/Mtok$1.14$0.35
Coding index42.04.9
Independent measurements behind it11
Cost per solved taskNot computableNot computable
SourceOpenRouter listingOpenRouter listing

Questions

Is DeepSeek V3 0324 better than Gemma 3 27B for coding?

Pick DeepSeek V3 0324 for coding: it scores 42.0 on the Vibe Coding Index against 4.9 for Gemma 3 27B. Pick Gemma 3 27B when output tokens cost 3.3x less, or when the benchmark that matters is Creative writing (Lech Mazur), where it leads.

Which is cheaper, DeepSeek V3 0324 or Gemma 3 27B?

DeepSeek V3 0324 costs $0.29 per million input tokens and $1.14 per million output tokens; Gemma 3 27B costs $0.08 and $0.35. Cost per solved task is known for neither of them.

Which has the bigger context window, DeepSeek V3 0324 or Gemma 3 27B?

DeepSeek V3 0324, with 164K tokens against 131K.

How many benchmarks is the DeepSeek V3 0324 vs Gemma 3 27B comparison based on?

21 benchmarks have a result for both models as of October 21, 2026. DeepSeek V3 0324's coding index rests on 1 independent benchmark and Gemma 3 27B's on 1. Where only one model has a result, the page says so instead of naming a winner.

Compare next