Enter opens the highlighted result. Esc closes.

R1 0528 vs Gemma 4 31B

Gemma 4 31B leads R1 0528 on the Artificial Analysis Intelligence Index, 14.7 to 13.1, and Gemma 4 31B costs 5.5x less per output token ($0.40 vs $2.18 per million). Updated October 22, 2026; 36 benchmarks have a result for both.

Replace R1 0528 Replace Gemma 4 31B
R1 0528DeepSeek
Coding
No data
Overall
46.8
Output $/M
$2.18
1
Gemma 4 31BGoogle
Coding
39.0
Overall
42.1
Output $/M
$0.40
35
vs boards won of 36 shared

R1 0528 or Gemma 4 31B: which to pick

Only Gemma 4 31B has a coding score, 39.0 on the Vibe Coding Index; R1 0528 has none, so the two cannot be ranked on coding.

Pick R1 0528 for work measured by GPQA Diamond (Epoch AI run), where it leads.

Model AR1 0528DeepSeekNot benchmarkedNo coding benchmarkNo cost per solved taskModel BGemma 4 31BGoogle39.0/1002 independent benchmarksNo cost per solved task
Shared results
36
R1 0528 only
27
Gemma 4 31B only
36
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
R1 0528 not benchmarked, Gemma 4 31B 39.0 from 2 benchmarks.
Widest category
Agentic work, 28.9 points apart on average over 1 shared benchmark.
Widest single gap
IFBench: R1 0528 39.6% against Gemma 4 31B 75.6%, 36.0 points apart on a 0 to 100 scale.
Output price
R1 0528 $2.18 against Gemma 4 31B $0.40 per million output tokens, 5.5x apart.
Context window
R1 0528 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
CategoryR1 0528BarsGemma 4 31BTies
Coding05
Intelligence01
Agentic work01
Reasoning15
Math03
Knowledge04
General016
All shared boards135
CodingLike-for-likeSix of these boards feed the Vibe Coding Index05
Coding benchmarks for R1 0528 and Gemma 4 31B
BenchmarkR1 0528Gemma 4 31BLeader
Aider polyglot71.4%Not measuredOnly R1 0528 measured
LiveBench codingNot measured49.0%Only Gemma 4 31B measured
LMArena WebDevNot measured1367Only Gemma 4 31B measured
AA Coding IndexNot measured43.4Only Gemma 4 31B measured
LiveCodeBench (Artificial Analysis run)77.0%Not measuredOnly R1 0528 measured
SciCodeNot measured45.5%Only Gemma 4 31B measured
Terminal-Bench Hard (Artificial Analysis run)15.9%36.4%Gemma 4 31B
WeirdML41.6%52.3%Gemma 4 31B
ALE-Bench804.1925.5Gemma 4 31B
LMArena Text: coding14271459Gemma 4 31B
LMArena Text: software and IT14351463Gemma 4 31B
LMArena WebDev: ReactNot measured1359Only Gemma 4 31B measured
LMArena WebDev: HTMLNot measured1380Only Gemma 4 31B measured
SciCodeNot measured8.9%Only Gemma 4 31B measured
Neither model has a result on the other 29 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
Intelligence benchmarks for R1 0528 and Gemma 4 31B
BenchmarkR1 0528Gemma 4 31BLeader
AA Intelligence Index13.114.7Gemma 4 31B
Epoch Capabilities IndexNot measured143.0Only Gemma 4 31B measured
Agentic workLike-for-likePublic boards, shown as published, kept out of the index01
Agentic work benchmarks for R1 0528 and Gemma 4 31B
BenchmarkR1 0528Gemma 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 Telecom36.5%65.5%Gemma 4 31B
METR time horizons53.8%Not measuredOnly R1 0528 measured
DeepResearch Bench35.1%Not measuredOnly R1 0528 measured
BrowseComp-Plus16.4%Not measuredOnly R1 0528 measured
Neither model has a result on the other 34 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index15
Reasoning benchmarks for R1 0528 and Gemma 4 31B
BenchmarkR1 0528Gemma 4 31BLeader
LiveBench reasoningNot measured68.3%Only Gemma 4 31B measured
GPQA Diamond (Artificial Analysis run)81.3%85.7%Gemma 4 31B
Humanity's Last Exam (Artificial Analysis run)15.8%23.6%Gemma 4 31B
GPQA Diamond (Epoch AI run)76.3%75.8%R1 0528
ARC-AGI21.2%Not measuredOnly R1 0528 measured
ARC-AGI-21.1%Not measuredOnly R1 0528 measured
SimpleBench40.8%Not measuredOnly R1 0528 measured
LMArena Text: hard prompts14151448Gemma 4 31B
LMArena Text: expert13941464Gemma 4 31B
LMArena Text: hard prompts in English14221456Gemma 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 10 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index03
Math benchmarks for R1 0528 and Gemma 4 31B
BenchmarkR1 0528Gemma 4 31BLeader
AA Math Index76.0Not measuredOnly R1 0528 measured
AIME 2025 (Artificial Analysis run)76.0%Not measuredOnly R1 0528 measured
OTIS Mock AIME66.4%73.3%Gemma 4 31B
MATH level 596.6%Not measuredOnly R1 0528 measured
LMArena Text: math14001463Gemma 4 31B
LMArena Text: mathematical work14151465Gemma 4 31B
MathArena12.9%Not measuredOnly R1 0528 measured
Neither model has a result on the other 7 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index04
Knowledge benchmarks for R1 0528 and Gemma 4 31B
BenchmarkR1 0528Gemma 4 31BLeader
CaseLaw (Vals AI)Not measured52.6%Only Gemma 4 31B measured
MortgageTax (Vals AI)Not measured61.4%Only Gemma 4 31B measured
MMLU Pro (Artificial Analysis run)84.9%Not measuredOnly R1 0528 measured
SimpleQA VerifiedNot measured10.4%Only Gemma 4 31B measured
LMArena Text: science14301457Gemma 4 31B
LMArena Text: medicine and healthcare14451451Gemma 4 31B
LMArena Text: legal and government14191438Gemma 4 31B
LMArena Text: business and finance14021430Gemma 4 31B
FACTS Benchmark SuiteNot measured45.3%Only Gemma 4 31B measured
PRBench Finance32.7%Not measuredOnly R1 0528 measured
PRBench Legal36.6%Not measuredOnly R1 0528 measured
Neither model has a result on the other 9 boards here
Web designDirectional onlyPublic boards, shown as published, kept out of the index8 boards, none shared
Web design benchmarks for R1 0528 and Gemma 4 31B
BenchmarkR1 0528Gemma 4 31BLeader
Design Arena: websites1149Not measuredOnly R1 0528 measured
Design Arena: UI components1100Not measuredOnly R1 0528 measured
Design Arena: game dev1097Not measuredOnly R1 0528 measured
Design Arena: data viz1182Not measuredOnly R1 0528 measured
Design Arena: 3D1110Not measuredOnly R1 0528 measured
Design Arena: all code categories1138Not measuredOnly R1 0528 measured
Design Arena: SVG1038Not measuredOnly R1 0528 measured
LiveBench designNot measured58.4%Only Gemma 4 31B measured
Neither model has a result on the other 1 board here
GeneralLike-for-likePublic boards, shown as published, kept out of the index016
General benchmarks for R1 0528 and Gemma 4 31B
BenchmarkR1 0528Gemma 4 31BLeader
LiveBench data analysisNot measured53.3%Only Gemma 4 31B measured
SAGE (Vals AI)Not measured55.0%Only Gemma 4 31B measured
IFBench39.6%75.6%Gemma 4 31B
Long context reasoning (AA-LCR)55.7%69.7%Gemma 4 31B
DTBenchNot measured82.7%Only Gemma 4 31B measured
LMCANot measured39.3%Only Gemma 4 31B measured
Fiction.LiveBench75.0%Not measuredOnly R1 0528 measured
Creative writing (Lech Mazur)81.9%Not measuredOnly R1 0528 measured
Surface Evolver BenchNot measured30.6%Only Gemma 4 31B measured
LMArena Text14281443Gemma 4 31B
LMArena Text: instruction following13821433Gemma 4 31B
LMArena Text: creative writing14051417Gemma 4 31B
LMArena VisionNot measured1276Only Gemma 4 31B measured
LMArena Text: English14371451Gemma 4 31B
LMArena Text: non-English14121430Gemma 4 31B
LMArena Text: Chinese14421475Gemma 4 31B
LMArena Text: German1404Not measuredOnly R1 0528 measured
LMArena Text: Japanese1391Not measuredOnly R1 0528 measured
LMArena Text: Korean1360Not measuredOnly R1 0528 measured
LMArena Text: French14171436Gemma 4 31B
LMArena Text: Spanish14111442Gemma 4 31B
LMArena Text: Russian14231460Gemma 4 31B
LMArena Text: Polish14281433Gemma 4 31B
LMArena Text: multi-turn14051452Gemma 4 31B
LMArena Text: longer queries13911446Gemma 4 31B
LMArena Text: writing and language14021432Gemma 4 31B
LMArena Text: entertainment and media14001405Gemma 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
EQ-Bench 4Not measured1120Only Gemma 4 31B measured
Neither model has a result on the other 12 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for R1 0528 and Gemma 4 31B
BenchmarkR1 0528Gemma 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

R1 0528$0.0016Fits in one request
Gemma 4 31B$0.0003Fits in one request

Gemma 4 31B costs less

Repository review

50K input + 3K output tokens

R1 0528$0.032Fits in one request
Gemma 4 31B$0.0082Fits in one request

Gemma 4 31B costs less

Agent loop

200K input + 10K output tokens

R1 0528$0.122Exceeds 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 R1 0528 and Gemma 4 31B
SpecificationDeepSeek R1 0528Google Gemma 4 31B
MakerDeepSeekGoogle
WeightsOpen weightsOpen weights
ReleasedMay 28, 2025April 2, 2026
Context window164K262K
Input $/Mtok$0.50$0.14
Output $/Mtok$2.18$0.40
Coding indexNot benchmarked39.0
Independent measurements behind it02
Cost per solved taskNot computableNot computable
SourceOpenRouter listingOpenRouter listing

Questions

Is R1 0528 better than Gemma 4 31B for coding?

Only Gemma 4 31B has a coding score, 39.0 on the Vibe Coding Index; R1 0528 has none, so the two cannot be ranked on coding. Pick R1 0528 for work measured by GPQA Diamond (Epoch AI run), where it leads.

Which is cheaper, R1 0528 or Gemma 4 31B?

R1 0528 costs $0.50 per million input tokens and $2.18 per million output tokens; Gemma 4 31B costs $0.14 and $0.40. Cost per solved task is known for neither of them.

Which has the bigger context window, R1 0528 or Gemma 4 31B?

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

How many benchmarks is the R1 0528 vs Gemma 4 31B comparison based on?

36 benchmarks have a result for both models as of October 22, 2026. R1 0528's coding index rests on 0 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