Enter opens the highlighted result. Esc closes.

Gemini 2.5 Flash vs Gemma 4 31B

Updated October 21, 2026. 46 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.

Replace Gemini 2.5 Flash Replace Gemma 4 31B
Gemini 2.5 FlashGoogle
Coding
28.7
Overall
15.5
Output $/M
$2.50
3
Gemma 4 31BGoogle
Coding
38.3
Overall
14.7
Output $/M
$0.40
43
vs boards won of 46 shared

Decision reading

Pick Gemma 4 31B for coding: it scores 38.3 on the Vibe Coding Index against 28.7 for Gemini 2.5 Flash.

Pick Gemini 2.5 Flash when the context window is larger, 1.05M against 262K tokens, or when what matters is AA Intelligence Index or LMArena Vision: captioning, where it leads.

Model AGemini 2.5 FlashGoogle28.7/1001 independent benchmark$0.466 per solved taskModel BGemma 4 31BGoogle38.3/1002 independent benchmarksNo cost per solved task
Shared results
46
Gemini 2.5 Flash only
23
Gemma 4 31B only
22
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
Gemini 2.5 Flash 28.7 from 1 benchmark, Gemma 4 31B 38.3 from 2 benchmarks.
Widest category
Agentic work, 33.9 points apart on average over 1 shared benchmark.
Widest single gap
τ²-Bench Telecom: Gemini 2.5 Flash 31.6% against Gemma 4 31B 65.5%, 33.9 points apart on a 0 to 100 scale.
Output price
Gemini 2.5 Flash $2.50 against Gemma 4 31B $0.40 per million output tokens, 6.3x apart.
Context window
Gemini 2.5 Flash takes 1.05M 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
CategoryGemini 2.5 FlashBarsGemma 4 31BTies
Coding03
Intelligence11
Agentic work01
Reasoning07
Math02
Knowledge04
General225
All shared boards343
CodingLike-for-likeFour of these boards feed the Vibe Coding Index03
Coding benchmarks for Gemini 2.5 Flash and Gemma 4 31B
BenchmarkGemini 2.5 FlashGemma 4 31BLeader
SWE-bench Verified28.7%Not measuredOnly Gemini 2.5 Flash 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)69.5%Not measuredOnly Gemini 2.5 Flash measured
SciCodeNot measured45.5%Only Gemma 4 31B measured
Terminal-Bench Hard (Artificial Analysis run)13.6%36.4%Gemma 4 31B
Terminal-Bench17.1%Not measuredOnly Gemini 2.5 Flash measured
WeirdMLNot measured52.3%Only Gemma 4 31B measured
ALE-BenchNot measured925.5Only Gemma 4 31B measured
LMArena Text: coding14231459Gemma 4 31B
LMArena Text: software and IT14251463Gemma 4 31B
LMArena WebDev: ReactNot measured1359Only Gemma 4 31B measured
LMArena WebDev: HTMLNot measured1380Only Gemma 4 31B measured
Neither model has a result on the other 24 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index11
Intelligence benchmarks for Gemini 2.5 Flash and Gemma 4 31B
BenchmarkGemini 2.5 FlashGemma 4 31BLeader
AA Intelligence Index15.514.7Gemini 2.5 Flash
Epoch Capabilities Index142.9143.0Gemma 4 31B
Agentic workLike-for-likePublic boards, shown as published, kept out of the index01
Agentic work benchmarks for Gemini 2.5 Flash and Gemma 4 31B
BenchmarkGemini 2.5 FlashGemma 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 Telecom31.6%65.5%Gemma 4 31B
BALROG33.5%Not measuredOnly Gemini 2.5 Flash measured
Vending-Bench 2$549Not measuredOnly Gemini 2.5 Flash measured
Neither model has a result on the other 32 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index07
Reasoning benchmarks for Gemini 2.5 Flash and Gemma 4 31B
BenchmarkGemini 2.5 FlashGemma 4 31BLeader
LiveBench reasoningNot measured68.3%Only Gemma 4 31B measured
GPQA Diamond (Artificial Analysis run)79.0%85.7%Gemma 4 31B
Humanity's Last Exam (Artificial Analysis run)12.1%23.6%Gemma 4 31B
GPQA Diamond (Epoch AI run)Not measured75.8%Only Gemma 4 31B measured
ARC-AGI33.3%Not measuredOnly Gemini 2.5 Flash measured
ARC-AGI-22.5%Not measuredOnly Gemini 2.5 Flash measured
SimpleBench41.2%Not measuredOnly Gemini 2.5 Flash measured
LMArena Text: hard prompts14221448Gemma 4 31B
LMArena Text: expert14261464Gemma 4 31B
LMArena Text: hard prompts in English14201456Gemma 4 31B
LMArena Vision: diagrams12451283Gemma 4 31B
LMArena Vision: homework12441310Gemma 4 31B
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 Gemini 2.5 Flash and Gemma 4 31B
BenchmarkGemini 2.5 FlashGemma 4 31BLeader
AA Math Index73.3Not measuredOnly Gemini 2.5 Flash measured
AIME 2025 (Artificial Analysis run)73.3%Not measuredOnly Gemini 2.5 Flash measured
OTIS Mock AIMENot measured73.3%Only Gemma 4 31B measured
LMArena Text: math14111463Gemma 4 31B
LMArena Text: mathematical work14221465Gemma 4 31B
Neither model has a result on the other 8 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index04
Knowledge benchmarks for Gemini 2.5 Flash and Gemma 4 31B
BenchmarkGemini 2.5 FlashGemma 4 31BLeader
CaseLaw (Vals AI)Not measured52.6%Only Gemma 4 31B measured
MortgageTax (Vals AI)Not measured61.4%Only Gemma 4 31B measured
MedCode (Vals AI)40.4%Not measuredOnly Gemini 2.5 Flash measured
MMLU Pro (Artificial Analysis run)83.2%Not measuredOnly Gemini 2.5 Flash measured
SimpleQA VerifiedNot measured10.4%Only Gemma 4 31B measured
LMArena Text: science14331457Gemma 4 31B
LMArena Text: medicine and healthcare14281451Gemma 4 31B
LMArena Text: legal and government14301438Gemma 4 31B
LMArena Text: business and finance14041430Gemma 4 31B
Neither model has a result on the other 8 boards here
Web designDirectional onlyPublic boards, shown as published, kept out of the index8 boards, none shared
Web design benchmarks for Gemini 2.5 Flash and Gemma 4 31B
BenchmarkGemini 2.5 FlashGemma 4 31BLeader
Design Arena: websites1074Not measuredOnly Gemini 2.5 Flash measured
Design Arena: UI components1005Not measuredOnly Gemini 2.5 Flash measured
Design Arena: game dev1005Not measuredOnly Gemini 2.5 Flash measured
Design Arena: data viz1060Not measuredOnly Gemini 2.5 Flash measured
Design Arena: 3D1021Not measuredOnly Gemini 2.5 Flash measured
Design Arena: all code categories1059Not measuredOnly Gemini 2.5 Flash measured
Design Arena: SVG1018Not measuredOnly Gemini 2.5 Flash 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 index225
General benchmarks for Gemini 2.5 Flash and Gemma 4 31B
BenchmarkGemini 2.5 FlashGemma 4 31BLeader
LiveBench data analysisNot measured53.3%Only Gemma 4 31B measured
SAGE (Vals AI)44.8%55.0%Gemma 4 31B
MedScribe (Vals AI)83.0%Not measuredOnly Gemini 2.5 Flash measured
IFBench50.3%75.6%Gemma 4 31B
Long context reasoning (AA-LCR)65.3%69.7%Gemma 4 31B
DTBench76.5%82.7%Gemma 4 31B
LMCA27.5%39.3%Gemma 4 31B
Surface Evolver BenchNot measured30.6%Only Gemma 4 31B measured
LMArena Text14171443Gemma 4 31B
LMArena Text: instruction following14051433Gemma 4 31B
LMArena Text: creative writing14001417Gemma 4 31B
LMArena Vision12351276Gemma 4 31B
LMArena Text: English14181451Gemma 4 31B
LMArena Text: non-English14091430Gemma 4 31B
LMArena Text: Chinese14491475Gemma 4 31B
LMArena Text: German1417Not measuredOnly Gemini 2.5 Flash measured
LMArena Text: Japanese1396Not measuredOnly Gemini 2.5 Flash measured
LMArena Text: Korean1385Not measuredOnly Gemini 2.5 Flash measured
LMArena Text: French14321436Gemma 4 31B
LMArena Text: Spanish14221442Gemma 4 31B
LMArena Text: Russian14151460Gemma 4 31B
LMArena Text: Polish14131433Gemma 4 31B
LMArena Text: multi-turn14081452Gemma 4 31B
LMArena Text: longer queries14191446Gemma 4 31B
LMArena Text: writing and language14051432Gemma 4 31B
LMArena Text: entertainment and media13931405Gemma 4 31B
LMArena Vision: English12391271Gemma 4 31B
LMArena Vision: Chinese12841326Gemma 4 31B
LMArena Vision: reading text12371282Gemma 4 31B
LMArena Vision: humor12251280Gemma 4 31B
LMArena Vision: captioning12771218Gemini 2.5 Flash
LMArena Vision: recognizing things12521226Gemini 2.5 Flash
LMArena Vision: creative writing12411285Gemma 4 31B
LMArena DocumentNot measured1425Only 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 Gemini 2.5 Flash and Gemma 4 31B
BenchmarkGemini 2.5 FlashGemma 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

Gemini 2.5 Flash$0.0015Fits in one request
Gemma 4 31B$0.0004Fits in one request

Gemma 4 31B costs less

Repository review

50K input + 3K output tokens

Gemini 2.5 Flash$0.022Fits in one request
Gemma 4 31B$0.011Fits in one request

Gemma 4 31B costs less

Agent loop

200K input + 10K output tokens

Gemini 2.5 Flash$0.085Fits in one request
Gemma 4 31B$0.044Fits 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 Gemini 2.5 Flash and Gemma 4 31B
SpecificationGoogle Gemini 2.5 FlashGoogle Gemma 4 31B
MakerGoogleGoogle
WeightsNo public weights listedOpen weights
ReleasedJune 17, 2025April 2, 2026
Context window1.05M262K
Input $/Mtok$0.30$0.20
Output $/Mtok$2.50$0.40
Coding index28.738.3
Independent measurements behind it12
Cost per solved task$0.466Not computable
SourceOpenRouter listingOpenRouter listing

Questions

Is Gemini 2.5 Flash better than Gemma 4 31B for coding?

Pick Gemma 4 31B for coding: it scores 38.3 on the Vibe Coding Index against 28.7 for Gemini 2.5 Flash. Pick Gemini 2.5 Flash when the context window is larger, 1.05M against 262K tokens, or when what matters is AA Intelligence Index or LMArena Vision: captioning, where it leads.

Which is cheaper, Gemini 2.5 Flash or Gemma 4 31B?

Gemini 2.5 Flash costs $0.30 per million input tokens and $2.50 per million output tokens; Gemma 4 31B costs $0.20 and $0.40. Cost per solved task is known for Gemini 2.5 Flash only.

Which has the bigger context window, Gemini 2.5 Flash or Gemma 4 31B?

Gemini 2.5 Flash, with 1.05M tokens against 262K.

How many benchmarks is the Gemini 2.5 Flash vs Gemma 4 31B comparison based on?

46 benchmarks have a result for both models as of October 21, 2026. Gemini 2.5 Flash's coding index rests on 1 independent benchmark 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