LLMBenchmark.ai

Enter opens the highlighted result. Esc closes.

GPT-5.4 vs Muse Spark 1.1

Updated October 21, 2026. 82 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.

Replace GPT-5.4 Replace Muse Spark 1.1
GPT-5.4OpenAI
Coding
37.7
Overall
39.0
Output $/M
$15.00
37
Muse Spark 1.1Meta
Coding
48.9
Overall
33.7
Output $/M
$4.25
44
vs boards won of 82 shared, 1 tied

Decision reading

Pick Muse Spark 1.1 for coding: it scores 48.9 on the Vibe Coding Index against 37.7 for GPT-5.4.

Pick GPT-5.4 when the context window is larger, 1.05M against 1.05M tokens, or when what matters is Code migration (Vals AI) or ProgramBench (Vals AI), where it leads.

Model AGPT-5.4OpenAI37.7/1001 independent benchmark$1.02 per solved taskModel BMuse Spark 1.1Meta48.9/1001 independent benchmark$0.43 per solved task
Shared results
82
GPT-5.4 only
27
Muse Spark 1.1 only
14
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
GPT-5.4 37.7 from 1 benchmark, Muse Spark 1.1 48.9 from 1 benchmark.
Widest category
Agentic work, 14.0 points apart on average over 7 shared benchmarks.
Widest single gap
Design Arena: 3D: GPT-5.4 1065 against Muse Spark 1.1 1260, 24.4 points apart on a 0 to 100 scale.
Output price
GPT-5.4 $15.00 against Muse Spark 1.1 $4.25 per million output tokens, 3.5x apart.
Context window
GPT-5.4 takes 1.05M tokens, Muse Spark 1.1 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
CategoryGPT-5.4BarsMuse Spark 1.1Ties
Coding211
Intelligence20
Agentic work25
Reasoning73
Math30
Knowledge46
Web design08
General17111 tie
All shared boards37441 tie
CodingLike-for-likeFour of these boards feed the Vibe Coding Index211
Coding benchmarks for GPT-5.4 and Muse Spark 1.1
BenchmarkGPT-5.4Muse Spark 1.1Leader
LMArena WebDev14641542Muse Spark 1.1
AA Coding Index71.171.3Muse Spark 1.1
LiveCodeBench (Vals AI)84.1%85.9%Muse Spark 1.1
SWE-bench Verified (Vals AI)78.2%82.0%Muse Spark 1.1
Vibe Code Bench (Vals AI)67.4%72.2%Muse Spark 1.1
Code migration (Vals AI)35.0%31.1%GPT-5.4
ProgramBench (Vals AI)0.5%0.0%GPT-5.4
Terminal-Bench 2.1 (Vals AI)Not measured69.3%Only Muse Spark 1.1 measured
SkillsBench (Vals AI)51.7%59.2%Muse Spark 1.1
SciCodeNot measured58.8%Only Muse Spark 1.1 measured
Terminal-Bench Hard (Artificial Analysis run)57.6%Not measuredOnly GPT-5.4 measured
Terminal-Bench81.8%Not measuredOnly GPT-5.4 measured
GSO-Bench31.4%Not measuredOnly GPT-5.4 measured
WeirdML77.7%Not measuredOnly GPT-5.4 measured
MirrorCode15.6%Not measuredOnly GPT-5.4 measured
ALE-Bench1607.0Not measuredOnly GPT-5.4 measured
LMArena Text: coding14961498Muse Spark 1.1
LMArena Text: software and IT14911498Muse Spark 1.1
LMArena WebDev: React14541532Muse Spark 1.1
LMArena WebDev: HTML14781540Muse Spark 1.1
LMArena WebDev: from an image14361540Muse Spark 1.1
Neither model has a result on the other 17 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
Intelligence benchmarks for GPT-5.4 and Muse Spark 1.1
BenchmarkGPT-5.4Muse Spark 1.1Leader
AA Intelligence Index39.033.7GPT-5.4
Epoch Capabilities Index157.3154.1GPT-5.4
Agentic workLike-for-likePublic boards, shown as published, kept out of the index25
Agentic work benchmarks for GPT-5.4 and Muse Spark 1.1
BenchmarkGPT-5.4Muse Spark 1.1Leader
AA Agentic IndexNot measured25.8Only Muse Spark 1.1 measured
Design Arena: agents building web apps10591207Muse Spark 1.1
Design Arena: agents building full-stack apps10171199Muse Spark 1.1
Design Arena: agents building mobile apps10871177Muse Spark 1.1
Design Arena: agents building Android apps10371184Muse Spark 1.1
Design Arena: agents building Godot games11351103GPT-5.4
Design Arena: agents building gamesNot measured1183Only Muse Spark 1.1 measured
Design Arena: HTML slidesNot measured1194Only Muse Spark 1.1 measured
Design Arena: PowerPoint slidesNot measured1156Only Muse Spark 1.1 measured
Design Arena: spreadsheetsNot measured974Only Muse Spark 1.1 measured
Finance Agent (Vals AI)Not measured57.2%Only Muse Spark 1.1 measured
Legal Research Bench (Vals AI)Not measured38.0%Only Muse Spark 1.1 measured
Excel Modeling Benchmark (Vals AI)Not measured56.4%Only Muse Spark 1.1 measured
Time Horizon Index (Vals AI)Not measured1.7%Only Muse Spark 1.1 measured
τ²-Bench Telecom87.1%Not measuredOnly GPT-5.4 measured
APEX-Agents52.4%31.8%GPT-5.4
METR time horizons74.3%Not measuredOnly GPT-5.4 measured
DeepResearch Bench35.1%Not measuredOnly GPT-5.4 measured
PostTrainBench19.0%Not measuredOnly GPT-5.4 measured
Vending-Bench 2$6,144$6,520Muse Spark 1.1
Neither model has a result on the other 19 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index73
Reasoning benchmarks for GPT-5.4 and Muse Spark 1.1
BenchmarkGPT-5.4Muse Spark 1.1Leader
GPQA Diamond91.7%91.2%GPT-5.4
MMMU Pro (Vals AI)87.5%86.6%GPT-5.4
GPQA Diamond (Artificial Analysis run)92.0%89.8%GPT-5.4
Humanity's Last Exam (Artificial Analysis run)43.7%46.2%Muse Spark 1.1
GPQA Diamond (Epoch AI run)93.3%Not measuredOnly GPT-5.4 measured
ARC-AGI93.7%Not measuredOnly GPT-5.4 measured
ARC-AGI-274.0%Not measuredOnly GPT-5.4 measured
Chess puzzles48.0%34.0%GPT-5.4
Mystery game puzzles37.0%Not measuredOnly GPT-5.4 measured
EBR-bench25.4%Not measuredOnly GPT-5.4 measured
LMArena Text: hard prompts14851485Muse Spark 1.1
LMArena Text: expert15071478GPT-5.4
LMArena Text: hard prompts in English14791480Muse Spark 1.1
LMArena Vision: diagrams13271305GPT-5.4
LMArena Vision: homework13361287GPT-5.4
Neither model has a result on the other 6 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index30
Math benchmarks for GPT-5.4 and Muse Spark 1.1
BenchmarkGPT-5.4Muse Spark 1.1Leader
AIME96.7%Not measuredOnly GPT-5.4 measured
OTIS Mock AIME97.8%Not measuredOnly GPT-5.4 measured
FrontierMath, tiers 1 to 378.6%Not measuredOnly GPT-5.4 measured
FrontierMath, tier 449.0%Not measuredOnly GPT-5.4 measured
ProofBench56.0%39.0%GPT-5.4
LMArena Text: math14881482GPT-5.4
LMArena Text: mathematical work14971487GPT-5.4
Neither model has a result on the other 6 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index46
Knowledge benchmarks for GPT-5.4 and Muse Spark 1.1
BenchmarkGPT-5.4Muse Spark 1.1Leader
MMLU Pro87.5%88.7%Muse Spark 1.1
LegalBench (Vals AI)86.0%85.0%GPT-5.4
CaseLaw (Vals AI)63.8%Not measuredOnly GPT-5.4 measured
CorpFin (Vals AI)65.3%71.3%Muse Spark 1.1
TaxEval (Vals AI)74.0%79.7%Muse Spark 1.1
MortgageTax (Vals AI)68.3%66.1%GPT-5.4
MedQA (Vals AI)96.1%Not measuredOnly GPT-5.4 measured
MedCode (Vals AI)41.3%Not measuredOnly GPT-5.4 measured
SimpleQA Verified45.1%57.8%Muse Spark 1.1
LMArena Text: science14791484Muse Spark 1.1
LMArena Text: medicine and healthcare14581479Muse Spark 1.1
LMArena Text: legal and government14861481GPT-5.4
LMArena Text: business and finance14741470GPT-5.4
Neither model has a result on the other 4 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index08
Web design benchmarks for GPT-5.4 and Muse Spark 1.1
BenchmarkGPT-5.4Muse Spark 1.1Leader
Design Arena: websites11991270Muse Spark 1.1
Design Arena: UI components12151290Muse Spark 1.1
Design Arena: game dev11951269Muse Spark 1.1
Design Arena: data viz11961268Muse Spark 1.1
Design Arena: 3D10651260Muse Spark 1.1
Design Arena: all code categories11801276Muse Spark 1.1
Design Arena: SVG11411234Muse Spark 1.1
Design Arena: ASCII art11881270Muse Spark 1.1
Neither model has a result on the other 1 board here
GeneralLike-for-likePublic boards, shown as published, kept out of the index1711
General benchmarks for GPT-5.4 and Muse Spark 1.1
BenchmarkGPT-5.4Muse Spark 1.1Leader
Public Benefits Bench (Vals AI)Not measured61.3%Only Muse Spark 1.1 measured
SAGE (Vals AI)43.3%45.6%Muse Spark 1.1
MedScribe (Vals AI)77.5%88.9%Muse Spark 1.1
Vals Multimodal IndexNot measured66.7%Only Muse Spark 1.1 measured
IFBench73.9%Not measuredOnly GPT-5.4 measured
Long context reasoning (AA-LCR)82.0%77.7%GPT-5.4
DTBench94.4%94.4%Tie
LMCA52.0%49.9%GPT-5.4
CL-bench27.9%Not measuredOnly GPT-5.4 measured
Surface Evolver BenchNot measured52.5%Only Muse Spark 1.1 measured
Furniture assembly37.5%Not measuredOnly GPT-5.4 measured
CL-bench Life21.7%Not measuredOnly GPT-5.4 measured
LMArena Text14691478Muse Spark 1.1
LMArena Text: instruction following14691457GPT-5.4
LMArena Text: creative writing14391435GPT-5.4
LMArena Vision13031293GPT-5.4
LMArena Text: English14671478Muse Spark 1.1
LMArena Text: non-English14651471Muse Spark 1.1
LMArena Text: Chinese15201515GPT-5.4
LMArena Text: German14721469GPT-5.4
LMArena Text: Japanese14851450GPT-5.4
LMArena Text: Korean14491459Muse Spark 1.1
LMArena Text: French14931492GPT-5.4
LMArena Text: Spanish14521464Muse Spark 1.1
LMArena Text: Russian14801482Muse Spark 1.1
LMArena Text: Polish14721499Muse Spark 1.1
LMArena Text: multi-turn14821484Muse Spark 1.1
LMArena Text: longer queries14731462GPT-5.4
LMArena Text: writing and language14581448GPT-5.4
LMArena Text: entertainment and media14381438GPT-5.4
LMArena Vision: English13011292GPT-5.4
LMArena Vision: Chinese13541348GPT-5.4
LMArena Vision: reading text13121301GPT-5.4
LMArena Vision: humor12841288Muse Spark 1.1
LMArena Vision: recognizing things1251Not measuredOnly GPT-5.4 measured
LMArena Vision: creative writing12831253GPT-5.4
LMArena Document14711465GPT-5.4
Neither model has a result on the other 9 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for GPT-5.4 and Muse Spark 1.1
BenchmarkGPT-5.4Muse Spark 1.1Leader
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

GPT-5.4$0.010Fits in one request
Muse Spark 1.1$0.0034Fits in one request

Muse Spark 1.1 costs less

Repository review

50K input + 3K output tokens

GPT-5.4$0.170Fits in one request
Muse Spark 1.1$0.075Fits in one request

Muse Spark 1.1 costs less

Agent loop

200K input + 10K output tokens

GPT-5.4$0.650Fits in one request
Muse Spark 1.1$0.292Fits in one request

Muse Spark 1.1 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 GPT-5.4 and Muse Spark 1.1
SpecificationOpenAI GPT-5.4Meta Muse Spark 1.1
MakerOpenAIMeta
WeightsNo public weights listedNo public weights listed
ReleasedMarch 5, 2026July 16, 2026
Context window1.05M1.05M
Input $/Mtok$2.50$1.25
Output $/Mtok$15.00$4.25
Coding index37.748.9
Independent measurements behind it11
Cost per solved task$1.02$0.43
SourceOpenRouter listingOpenRouter listing

Questions

Is GPT-5.4 better than Muse Spark 1.1 for coding?

Pick Muse Spark 1.1 for coding: it scores 48.9 on the Vibe Coding Index against 37.7 for GPT-5.4. Pick GPT-5.4 when the context window is larger, 1.05M against 1.05M tokens, or when what matters is Code migration (Vals AI) or ProgramBench (Vals AI), where it leads.

Which is cheaper, GPT-5.4 or Muse Spark 1.1?

GPT-5.4 costs $2.50 per million input tokens and $15.00 per million output tokens; Muse Spark 1.1 costs $1.25 and $4.25. Per solved SWE-bench Verified task, GPT-5.4 costs $1.02 and Muse Spark 1.1 $0.43.

Which has the bigger context window, GPT-5.4 or Muse Spark 1.1?

GPT-5.4, with 1.05M tokens against 1.05M.

How many benchmarks is the GPT-5.4 vs Muse Spark 1.1 comparison based on?

82 benchmarks have a result for both models as of October 21, 2026. GPT-5.4's coding index rests on 1 independent benchmark and Muse Spark 1.1's on 1. Where only one model has a result, the page says so instead of naming a winner.

Compare next