Enter opens the highlighted result. Esc closes.

DeepSeek V3.2 vs Gemma 4 31B

Gemma 4 31B leads DeepSeek V3.2 on coding, 39.0 to 30.0 on the Vibe Coding Index, and Gemma 4 31B costs 2.2x less per output token ($0.40 vs $0.88 per million). Updated October 21, 2026; 40 benchmarks have a result for both.

Replace DeepSeek V3.2 Replace Gemma 4 31B
DeepSeek V3.2DeepSeek
Coding
30.0
Overall
37.9
Output $/M
$0.88
8
Gemma 4 31BGoogle
Coding
39.0
Overall
42.1
Output $/M
$0.40
32
vs boards won of 40 shared

DeepSeek V3.2 or Gemma 4 31B: which to pick

Pick Gemma 4 31B for coding: it scores 39.0 on the Vibe Coding Index against 30.0 for DeepSeek V3.2, though on 2 benchmarks against 3.

Pick DeepSeek V3.2 for work measured by AA Coding Index or AA Intelligence Index, where it leads.

Model ADeepSeek V3.2DeepSeek30.0/1003 independent benchmarks$0.99 per solved taskModel BGemma 4 31BGoogle39.0/1002 independent benchmarksNo cost per solved task
Shared results
40
DeepSeek V3.2 only
35
Gemma 4 31B only
31
Like-for-like categories
7 of 9

Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.

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 V3.2 30.0 from 3 benchmarks, Gemma 4 31B 39.0 from 2 benchmarks.
Widest category
Agentic work, 25.1 points apart on average over 1 shared benchmark.
Widest single gap
τ²-Bench Telecom: DeepSeek V3.2 90.6% against Gemma 4 31B 65.5%, 25.1 points apart on a 0 to 100 scale.
Output price
DeepSeek V3.2 $0.88 against Gemma 4 31B $0.40 per million output tokens, 2.2x apart.
Context window
DeepSeek V3.2 takes 164K tokens, Gemma 4 31B 262K.

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 V3.2BarsGemma 4 31BTies
Coding16
Intelligence20
Agentic work10
Reasoning14
Math02
Knowledge14
General216
All shared boards832
CodingLike-for-likeSix of these boards feed the Vibe Coding Index16
Coding benchmarks for DeepSeek V3.2 and Gemma 4 31B
BenchmarkDeepSeek V3.2Gemma 4 31BLeader
SWE-bench Verified70.0%Not measuredOnly DeepSeek V3.2 measured
SWE-bench Multilingual59.0%Not measuredOnly DeepSeek V3.2 measured
LiveBench codingNot measured49.0%Only Gemma 4 31B measured
LMArena WebDev13621367Gemma 4 31B
AA Coding Index44.243.4DeepSeek V3.2
LiveCodeBench (Vals AI)80.7%Not measuredOnly DeepSeek V3.2 measured
SWE-bench Verified (Vals AI)67.6%Not measuredOnly DeepSeek V3.2 measured
Vibe Code Bench (Vals AI)5.1%Not measuredOnly DeepSeek V3.2 measured
LiveCodeBench (Artificial Analysis run)86.2%Not measuredOnly DeepSeek V3.2 measured
SciCodeNot measured45.5%Only Gemma 4 31B measured
Terminal-Bench Hard (Artificial Analysis run)35.6%36.4%Gemma 4 31B
Terminal-Bench39.6%Not measuredOnly DeepSeek V3.2 measured
WeirdMLNot measured52.3%Only Gemma 4 31B measured
ALE-BenchNot measured925.5Only Gemma 4 31B measured
LMArena Text: coding14531459Gemma 4 31B
LMArena Text: software and IT14491463Gemma 4 31B
LMArena WebDev: React13501359Gemma 4 31B
LMArena WebDev: HTML13711380Gemma 4 31B
SciCodeNot measured8.9%Only Gemma 4 31B measured
Neither model has a result on the other 20 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
Intelligence benchmarks for DeepSeek V3.2 and Gemma 4 31B
BenchmarkDeepSeek V3.2Gemma 4 31BLeader
AA Intelligence Index21.514.7DeepSeek V3.2
Epoch Capabilities Index146.1143.0DeepSeek V3.2
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
Agentic work benchmarks for DeepSeek V3.2 and Gemma 4 31B
BenchmarkDeepSeek V3.2Gemma 4 31BLeader
AA Agentic IndexNot measured4.2Only Gemma 4 31B measured
LiveBench researchNot measured44.4%Only Gemma 4 31B measured
LiveBench browserNot measured4.0%Only Gemma 4 31B measured
LiveBench long agenticNot measured13.3%Only Gemma 4 31B measured
τ²-Bench Telecom90.6%65.5%DeepSeek V3.2
APEX-Agents21.3%Not measuredOnly DeepSeek V3.2 measured
Vending-Bench 2$1,034Not measuredOnly DeepSeek V3.2 measured
Neither model has a result on the other 34 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index14
Reasoning benchmarks for DeepSeek V3.2 and Gemma 4 31B
BenchmarkDeepSeek V3.2Gemma 4 31BLeader
GPQA Diamond80.3%Not measuredOnly DeepSeek V3.2 measured
LiveBench reasoningNot measured68.3%Only Gemma 4 31B measured
GPQA Diamond (Artificial Analysis run)84.0%85.7%Gemma 4 31B
Humanity's Last Exam (Artificial Analysis run)24.6%23.6%DeepSeek V3.2
GPQA Diamond (Epoch AI run)Not measured75.8%Only Gemma 4 31B measured
ARC-AGI57.0%Not measuredOnly DeepSeek V3.2 measured
ARC-AGI-24.0%Not measuredOnly DeepSeek V3.2 measured
LMArena Text: hard prompts14341448Gemma 4 31B
LMArena Text: expert14361464Gemma 4 31B
LMArena Text: hard prompts in English14461456Gemma 4 31B
LMArena Vision: diagramsNot measured1283Only Gemma 4 31B measured
LMArena Vision: homeworkNot measured1310Only Gemma 4 31B measured
Kagi LLM BenchmarkNot measured63.5%Only Gemma 4 31B measured
Neither model has a result on the other 9 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index02
Math benchmarks for DeepSeek V3.2 and Gemma 4 31B
BenchmarkDeepSeek V3.2Gemma 4 31BLeader
AIME84.6%Not measuredOnly DeepSeek V3.2 measured
MGSM (Vals AI)86.0%Not measuredOnly DeepSeek V3.2 measured
AA Math Index92.0Not measuredOnly DeepSeek V3.2 measured
AIME 2025 (Artificial Analysis run)92.0%Not measuredOnly DeepSeek V3.2 measured
OTIS Mock AIMENot measured73.3%Only Gemma 4 31B measured
LMArena Text: math14351463Gemma 4 31B
LMArena Text: mathematical work14401465Gemma 4 31B
Neither model has a result on the other 6 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index14
Knowledge benchmarks for DeepSeek V3.2 and Gemma 4 31B
BenchmarkDeepSeek V3.2Gemma 4 31BLeader
MMLU Pro84.9%Not measuredOnly DeepSeek V3.2 measured
LegalBench (Vals AI)76.1%Not measuredOnly DeepSeek V3.2 measured
CaseLaw (Vals AI)55.4%52.6%DeepSeek V3.2
CorpFin (Vals AI)51.0%Not measuredOnly DeepSeek V3.2 measured
TaxEval (Vals AI)68.2%Not measuredOnly DeepSeek V3.2 measured
MortgageTax (Vals AI)Not measured61.4%Only Gemma 4 31B measured
MedQA (Vals AI)93.9%Not measuredOnly DeepSeek V3.2 measured
MMLU Pro (Artificial Analysis run)86.2%Not measuredOnly DeepSeek V3.2 measured
SimpleQA VerifiedNot measured10.4%Only Gemma 4 31B measured
LMArena Text: science14461457Gemma 4 31B
LMArena Text: medicine and healthcare14401451Gemma 4 31B
LMArena Text: legal and government14251438Gemma 4 31B
LMArena Text: business and finance14101430Gemma 4 31B
FACTS Benchmark SuiteNot measured45.3%Only Gemma 4 31B measured
Neither model has a result on the other 4 boards here
Web designDirectional onlyPublic boards, shown as published, kept out of the index9 boards, none shared
Web design benchmarks for DeepSeek V3.2 and Gemma 4 31B
BenchmarkDeepSeek V3.2Gemma 4 31BLeader
Design Arena: websites1175Not measuredOnly DeepSeek V3.2 measured
Design Arena: UI components1149Not measuredOnly DeepSeek V3.2 measured
Design Arena: game dev1132Not measuredOnly DeepSeek V3.2 measured
Design Arena: data viz1158Not measuredOnly DeepSeek V3.2 measured
Design Arena: 3D1133Not measuredOnly DeepSeek V3.2 measured
Design Arena: all code categories1164Not measuredOnly DeepSeek V3.2 measured
Design Arena: SVG1032Not measuredOnly DeepSeek V3.2 measured
Design Arena: ASCII art1083Not measuredOnly DeepSeek V3.2 measured
LiveBench designNot measured58.4%Only Gemma 4 31B measured
GeneralLike-for-likePublic boards, shown as published, kept out of the index216
General benchmarks for DeepSeek V3.2 and Gemma 4 31B
BenchmarkDeepSeek V3.2Gemma 4 31BLeader
LiveBench data analysisNot measured53.3%Only Gemma 4 31B measured
SAGE (Vals AI)Not measured55.0%Only Gemma 4 31B measured
IFBench60.7%75.6%Gemma 4 31B
Long context reasoning (AA-LCR)73.3%69.7%DeepSeek V3.2
DTBench85.6%82.7%DeepSeek V3.2
LMCA28.8%39.3%Gemma 4 31B
CL-bench12.4%Not measuredOnly DeepSeek V3.2 measured
Surface Evolver BenchNot measured30.6%Only Gemma 4 31B measured
CL-bench Life7.4%Not measuredOnly DeepSeek V3.2 measured
LMArena Text14251443Gemma 4 31B
LMArena Text: instruction following14131433Gemma 4 31B
LMArena Text: creative writing14001417Gemma 4 31B
LMArena VisionNot measured1276Only Gemma 4 31B measured
LMArena Text: English14391451Gemma 4 31B
LMArena Text: non-English14091430Gemma 4 31B
LMArena Text: Chinese14551475Gemma 4 31B
LMArena Text: German1413Not measuredOnly DeepSeek V3.2 measured
LMArena Text: Japanese1374Not measuredOnly DeepSeek V3.2 measured
LMArena Text: Korean1370Not measuredOnly DeepSeek V3.2 measured
LMArena Text: French14291436Gemma 4 31B
LMArena Text: Spanish14241442Gemma 4 31B
LMArena Text: Russian14241460Gemma 4 31B
LMArena Text: Polish14191433Gemma 4 31B
LMArena Text: multi-turn14271452Gemma 4 31B
LMArena Text: longer queries14281446Gemma 4 31B
LMArena Text: writing and language14071432Gemma 4 31B
LMArena Text: entertainment and media13951405Gemma 4 31B
LMArena Vision: EnglishNot measured1271Only Gemma 4 31B measured
LMArena Vision: ChineseNot measured1326Only Gemma 4 31B measured
LMArena Vision: reading textNot measured1282Only Gemma 4 31B measured
LMArena Vision: humorNot measured1280Only Gemma 4 31B measured
LMArena Vision: captioningNot measured1218Only Gemma 4 31B measured
LMArena Vision: recognizing thingsNot measured1226Only Gemma 4 31B measured
LMArena Vision: creative writingNot measured1285Only Gemma 4 31B measured
LMArena DocumentNot measured1425Only Gemma 4 31B measured
Neither model has a result on the other 11 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for DeepSeek V3.2 and Gemma 4 31B
BenchmarkDeepSeek V3.2Gemma 4 31BLeader
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 V3.2$0.0007Fits in one request
Gemma 4 31B$0.0003Fits in one request

Gemma 4 31B costs less

Repository review

50K input + 3K output tokens

DeepSeek V3.2$0.017Fits in one request
Gemma 4 31B$0.0082Fits in one request

Gemma 4 31B costs less

Agent loop

200K input + 10K output tokens

DeepSeek V3.2$0.067Exceeds the context window
Gemma 4 31B$0.032Fits in one request

Gemma 4 31B 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 V3.2 and Gemma 4 31B
SpecificationDeepSeek V3.2Google Gemma 4 31B
MakerDeepSeekGoogle
WeightsOpen weightsOpen weights
ReleasedDecember 1, 2025April 2, 2026
Context window164K262K
Input $/Mtok$0.29$0.14
Output $/Mtok$0.88$0.40
Coding index30.039.0
Independent measurements behind it32
Cost per solved task$0.99Not computable
SourceOpenRouter listingOpenRouter listing

Questions

Is DeepSeek V3.2 better than Gemma 4 31B for coding?

Pick Gemma 4 31B for coding: it scores 39.0 on the Vibe Coding Index against 30.0 for DeepSeek V3.2, though on 2 benchmarks against 3. Pick DeepSeek V3.2 for work measured by AA Coding Index or AA Intelligence Index, where it leads.

Which is cheaper, DeepSeek V3.2 or Gemma 4 31B?

DeepSeek V3.2 costs $0.29 per million input tokens and $0.88 per million output tokens; Gemma 4 31B costs $0.14 and $0.40. Cost per solved task is known for DeepSeek V3.2 only.

Which has the bigger context window, DeepSeek V3.2 or Gemma 4 31B?

Gemma 4 31B, with 262K tokens against 164K.

How many benchmarks is the DeepSeek V3.2 vs Gemma 4 31B comparison based on?

40 benchmarks have a result for both models as of October 21, 2026. DeepSeek V3.2's coding index rests on 3 independent benchmarks and Gemma 4 31B's on 2. Where only one model has a result, the page says so instead of naming a winner.

Compare next

Embed this comparison on your site

Free under CC BY 4.0, while the source link stays visible. Preview the frame · Terms and other embeds

<iframe src="https://llmbenchmark.ai/embed/compare/deepseek-v3-2-vs-gemma-4-31b-it" title="DeepSeek V3.2 vs Gemma 4 31B, LLMBenchmark.ai" width="100%" height="360" style="border:0;max-width:100%" loading="lazy"></iframe>
<p style="font:13px/1.4 sans-serif;margin:6px 0 0">Source: <a href="https://llmbenchmark.ai/compare/deepseek-v3-2-vs-gemma-4-31b-it">LLMBenchmark.ai, DeepSeek V3.2 vs Gemma 4 31B</a></p>