LLMBenchmark.ai

Enter opens the highlighted result. Esc closes.

Gemma 4 31B vs Muse Spark 1.2

Updated October 21, 2026. 52 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.

Replace Gemma 4 31B Replace Muse Spark 1.2
Gemma 4 31BGoogle
Coding
38.3
Overall
14.7
Output $/M
$0.40
1
Muse Spark 1.2Meta
Coding
65.1
Overall
39.6
Output $/M
$4.25
51
vs boards won of 52 shared

Decision reading

Pick Muse Spark 1.2 for coding: it scores 65.1 on the Vibe Coding Index against 38.3 for Gemma 4 31B.

Pick Gemma 4 31B when output tokens cost 10.6x less, or when the benchmark that matters is SAGE (Vals AI), where it leads.

Model AGemma 4 31BGoogle38.3/1002 independent benchmarksNo cost per solved taskModel BMuse Spark 1.2Meta65.1/1002 independent benchmarks$0.64 per solved task
Shared results
52
Gemma 4 31B only
16
Muse Spark 1.2 only
45
Like-for-like categories
8 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
Gemma 4 31B 38.3 from 2 benchmarks, Muse Spark 1.2 65.1 from 2 benchmarks.
Widest category
Agentic work, 25.6 points apart on average over 4 shared benchmarks.
Widest single gap
SimpleQA Verified: Gemma 4 31B 10.4% against Muse Spark 1.2 60.3%, 49.9 points apart on a 0 to 100 scale.
Output price
Gemma 4 31B $0.40 against Muse Spark 1.2 $4.25 per million output tokens, 10.6x apart.
Context window
Gemma 4 31B takes 262K tokens, Muse Spark 1.2 1.05M.

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
CategoryGemma 4 31BBarsMuse Spark 1.2Ties
Coding09
Intelligence02
Agentic work04
Reasoning07
Math02
Knowledge06
Web design01
General120
All shared boards151
CodingLike-for-likeFour of these boards feed the Vibe Coding Index09
Coding benchmarks for Gemma 4 31B and Muse Spark 1.2
BenchmarkGemma 4 31BMuse Spark 1.2Leader
LiveBench coding49.0%78.2%Muse Spark 1.2
LMArena WebDev13671534Muse Spark 1.2
AA Coding Index43.472.2Muse Spark 1.2
SWE-bench Verified (Vals AI)Not measured86.6%Only Muse Spark 1.2 measured
Terminal-Bench 4 (Vals AI)Not measured6.1%Only Muse Spark 1.2 measured
Vibe Code Bench (Vals AI)Not measured79.1%Only Muse Spark 1.2 measured
Code migration (Vals AI)Not measured29.9%Only Muse Spark 1.2 measured
ProgramBench (Vals AI)Not measured0.5%Only Muse Spark 1.2 measured
IOI (Vals AI)Not measured21.8%Only Muse Spark 1.2 measured
Terminal-Bench 2.1 (Vals AI)Not measured69.7%Only Muse Spark 1.2 measured
SkillsBench (Vals AI)Not measured53.0%Only Muse Spark 1.2 measured
CyberBench (Vals AI)Not measured69.5%Only Muse Spark 1.2 measured
SciCode45.5%57.4%Muse Spark 1.2
Terminal-Bench Hard (Artificial Analysis run)36.4%Not measuredOnly Gemma 4 31B measured
FrontierSWENot measured12.0%Only Muse Spark 1.2 measured
WeirdML52.3%60.3%Muse Spark 1.2
ALE-Bench925.5Not measuredOnly Gemma 4 31B measured
LMArena Text: coding14591505Muse Spark 1.2
LMArena Text: software and IT14631505Muse Spark 1.2
LMArena WebDev: React13591526Muse Spark 1.2
LMArena WebDev: HTML13801543Muse Spark 1.2
Neither model has a result on the other 17 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
Intelligence benchmarks for Gemma 4 31B and Muse Spark 1.2
BenchmarkGemma 4 31BMuse Spark 1.2Leader
AA Intelligence Index14.739.6Muse Spark 1.2
Epoch Capabilities Index143.0155.1Muse Spark 1.2
Agentic workLike-for-likePublic boards, shown as published, kept out of the index04
Agentic work benchmarks for Gemma 4 31B and Muse Spark 1.2
BenchmarkGemma 4 31BMuse Spark 1.2Leader
AA Agentic Index4.243.2Muse Spark 1.2
Design Arena: agents building web appsNot measured1222Only Muse Spark 1.2 measured
Design Arena: agents building full-stack appsNot measured1210Only Muse Spark 1.2 measured
Design Arena: agents building mobile appsNot measured1182Only Muse Spark 1.2 measured
Design Arena: agents building Android appsNot measured1228Only Muse Spark 1.2 measured
Design Arena: agents building gamesNot measured1179Only Muse Spark 1.2 measured
Design Arena: HTML slidesNot measured1135Only Muse Spark 1.2 measured
Design Arena: PowerPoint slidesNot measured1198Only Muse Spark 1.2 measured
LiveBench research44.4%67.7%Muse Spark 1.2
LiveBench browser4.0%36.0%Muse Spark 1.2
LiveBench long agentic13.3%21.3%Muse Spark 1.2
SREBench (Vals AI)Not measured0.0%Only Muse Spark 1.2 measured
Finance Agent (Vals AI)Not measured60.6%Only Muse Spark 1.2 measured
Tax Agent Bench (Vals AI)Not measured32.8%Only Muse Spark 1.2 measured
Legal Research Bench (Vals AI)Not measured43.8%Only Muse Spark 1.2 measured
Excel Modeling Benchmark (Vals AI)Not measured57.0%Only Muse Spark 1.2 measured
BioMysteryBench (Vals AI)Not measured64.8%Only Muse Spark 1.2 measured
τ²-Bench Telecom65.5%Not measuredOnly Gemma 4 31B measured
APEX-AgentsNot measured45.4%Only Muse Spark 1.2 measured
Neither model has a result on the other 20 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index07
Reasoning benchmarks for Gemma 4 31B and Muse Spark 1.2
BenchmarkGemma 4 31BMuse Spark 1.2Leader
LiveBench reasoning68.3%90.0%Muse Spark 1.2
MMMU Pro (Vals AI)Not measured86.1%Only Muse Spark 1.2 measured
GPQA Diamond (Artificial Analysis run)85.7%90.4%Muse Spark 1.2
Humanity's Last Exam (Artificial Analysis run)23.6%45.5%Muse Spark 1.2
GPQA Diamond (Epoch AI run)75.8%Not measuredOnly Gemma 4 31B measured
SimpleBenchNot measured74.5%Only Muse Spark 1.2 measured
Chess puzzlesNot measured27.0%Only Muse Spark 1.2 measured
LMArena Text: hard prompts14481489Muse Spark 1.2
LMArena Text: expert14641475Muse Spark 1.2
LMArena Text: hard prompts in English14561487Muse Spark 1.2
LMArena Vision: diagrams12831317Muse Spark 1.2
LMArena Vision: homework1310Not measuredOnly 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 Gemma 4 31B and Muse Spark 1.2
BenchmarkGemma 4 31BMuse Spark 1.2Leader
OTIS Mock AIME73.3%Not measuredOnly Gemma 4 31B measured
ProofBenchNot measured43.0%Only Muse Spark 1.2 measured
LMArena Text: math14631484Muse Spark 1.2
LMArena Text: mathematical work14651480Muse Spark 1.2
Neither model has a result on the other 9 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index06
Knowledge benchmarks for Gemma 4 31B and Muse Spark 1.2
BenchmarkGemma 4 31BMuse Spark 1.2Leader
MMLU ProNot measured88.3%Only Muse Spark 1.2 measured
LegalBench (Vals AI)Not measured85.3%Only Muse Spark 1.2 measured
CaseLaw (Vals AI)52.6%Not measuredOnly Gemma 4 31B measured
CorpFin (Vals AI)Not measured70.9%Only Muse Spark 1.2 measured
TaxEval (Vals AI)Not measured80.4%Only Muse Spark 1.2 measured
MortgageTax (Vals AI)61.4%65.4%Muse Spark 1.2
MedCode (Vals AI)Not measured49.4%Only Muse Spark 1.2 measured
SimpleQA Verified10.4%60.3%Muse Spark 1.2
LMArena Text: science14571500Muse Spark 1.2
LMArena Text: medicine and healthcare14511496Muse Spark 1.2
LMArena Text: legal and government14381517Muse Spark 1.2
LMArena Text: business and finance14301484Muse Spark 1.2
Neither model has a result on the other 5 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index01
Web design benchmarks for Gemma 4 31B and Muse Spark 1.2
BenchmarkGemma 4 31BMuse Spark 1.2Leader
Design Arena: websitesNot measured1313Only Muse Spark 1.2 measured
Design Arena: UI componentsNot measured1300Only Muse Spark 1.2 measured
Design Arena: game devNot measured1302Only Muse Spark 1.2 measured
Design Arena: data vizNot measured1334Only Muse Spark 1.2 measured
Design Arena: 3DNot measured1301Only Muse Spark 1.2 measured
Design Arena: all code categoriesNot measured1313Only Muse Spark 1.2 measured
Design Arena: SVGNot measured1285Only Muse Spark 1.2 measured
Design Arena: ASCII artNot measured1284Only Muse Spark 1.2 measured
LiveBench design58.4%67.1%Muse Spark 1.2
GeneralLike-for-likePublic boards, shown as published, kept out of the index120
General benchmarks for Gemma 4 31B and Muse Spark 1.2
BenchmarkGemma 4 31BMuse Spark 1.2Leader
LiveBench data analysis53.3%76.5%Muse Spark 1.2
Public Benefits Bench (Vals AI)Not measured68.5%Only Muse Spark 1.2 measured
SAGE (Vals AI)55.0%47.7%Gemma 4 31B
MedScribe (Vals AI)Not measured90.1%Only Muse Spark 1.2 measured
Vals Multimodal IndexNot measured69.8%Only Muse Spark 1.2 measured
IFBench75.6%Not measuredOnly Gemma 4 31B measured
Long context reasoning (AA-LCR)69.7%79.0%Muse Spark 1.2
DTBench82.7%94.7%Muse Spark 1.2
LMCA39.3%48.4%Muse Spark 1.2
Surface Evolver Bench30.6%Not measuredOnly Gemma 4 31B measured
LMArena Text14431482Muse Spark 1.2
LMArena Text: instruction following14331462Muse Spark 1.2
LMArena Text: creative writing14171436Muse Spark 1.2
LMArena Vision12761304Muse Spark 1.2
LMArena Text: English14511483Muse Spark 1.2
LMArena Text: non-English14301476Muse Spark 1.2
LMArena Text: Chinese14751509Muse Spark 1.2
LMArena Text: KoreanNot measured1462Only Muse Spark 1.2 measured
LMArena Text: French14361495Muse Spark 1.2
LMArena Text: Spanish14421484Muse Spark 1.2
LMArena Text: Russian14601483Muse Spark 1.2
LMArena Text: Polish1433Not measuredOnly Gemma 4 31B measured
LMArena Text: multi-turn14521502Muse Spark 1.2
LMArena Text: longer queries14461473Muse Spark 1.2
LMArena Text: writing and language14321453Muse Spark 1.2
LMArena Text: entertainment and media14051443Muse Spark 1.2
LMArena Vision: English12711291Muse Spark 1.2
LMArena Vision: Chinese1326Not measuredOnly Gemma 4 31B measured
LMArena Vision: reading text12821313Muse Spark 1.2
LMArena Vision: humor1280Not measuredOnly Gemma 4 31B measured
LMArena Vision: captioning1218Not measuredOnly Gemma 4 31B measured
LMArena Vision: recognizing things1226Not measuredOnly Gemma 4 31B measured
LMArena Vision: creative writing1285Not measuredOnly Gemma 4 31B measured
LMArena Document1425Not measuredOnly 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 Gemma 4 31B and Muse Spark 1.2
BenchmarkGemma 4 31BMuse Spark 1.2Leader
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

Gemma 4 31B$0.0004Fits in one request
Muse Spark 1.2$0.0034Fits in one request

Gemma 4 31B costs less

Repository review

50K input + 3K output tokens

Gemma 4 31B$0.011Fits in one request
Muse Spark 1.2$0.075Fits in one request

Gemma 4 31B costs less

Agent loop

200K input + 10K output tokens

Gemma 4 31B$0.044Fits in one request
Muse Spark 1.2$0.292Fits 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 Gemma 4 31B and Muse Spark 1.2
SpecificationGoogle Gemma 4 31BMeta Muse Spark 1.2
MakerGoogleMeta
WeightsOpen weightsNo public weights listed
ReleasedApril 2, 2026August 5, 2026
Context window262K1.05M
Input $/Mtok$0.20$1.25
Output $/Mtok$0.40$4.25
Coding index38.365.1
Independent measurements behind it22
Cost per solved taskNot computable$0.64
SourceOpenRouter listingOpenRouter listing

Questions

Is Gemma 4 31B better than Muse Spark 1.2 for coding?

Pick Muse Spark 1.2 for coding: it scores 65.1 on the Vibe Coding Index against 38.3 for Gemma 4 31B. Pick Gemma 4 31B when output tokens cost 10.6x less, or when the benchmark that matters is SAGE (Vals AI), where it leads.

Which is cheaper, Gemma 4 31B or Muse Spark 1.2?

Gemma 4 31B costs $0.20 per million input tokens and $0.40 per million output tokens; Muse Spark 1.2 costs $1.25 and $4.25. Cost per solved task is known for Muse Spark 1.2 only.

Which has the bigger context window, Gemma 4 31B or Muse Spark 1.2?

Muse Spark 1.2, with 1.05M tokens against 262K.

How many benchmarks is the Gemma 4 31B vs Muse Spark 1.2 comparison based on?

52 benchmarks have a result for both models as of October 21, 2026. Gemma 4 31B's coding index rests on 2 independent benchmarks and Muse Spark 1.2'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/gemma-4-31b-it-vs-meta-muse-spark-1-2" title="Gemma 4 31B vs Muse Spark 1.2, 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/gemma-4-31b-it-vs-meta-muse-spark-1-2">LLMBenchmark.ai, Gemma 4 31B vs Muse Spark 1.2</a></p>