Enter opens the highlighted result. Esc closes.

DeepSeek V4 Pro 0423 vs Muse Spark 1.2

Muse Spark 1.2 leads DeepSeek V4 Pro 0423 on coding, 55.7 to 38.7 on the Vibe Coding Index, and DeepSeek V4 Pro 0423 costs 1.4x less per output token ($2.976 vs $4.25 per million). Updated October 22, 2026; 66 benchmarks have a result for both.

Replace DeepSeek V4 Pro 0423 Replace Muse Spark 1.2
DeepSeek V4 Pro 0423DeepSeek
Coding
38.7
Overall
50.0
Output $/M
$2.976
2
Muse Spark 1.2Meta
Coding
55.7
Overall
60.4
Output $/M
$4.25
64
vs boards won of 66 shared

DeepSeek V4 Pro 0423 or Muse Spark 1.2: which to pick

Pick Muse Spark 1.2 for coding: it scores 55.7 on the Vibe Coding Index against 38.7 for DeepSeek V4 Pro 0423.

Pick DeepSeek V4 Pro 0423 for output tokens 1.4x cheaper ($2.976 vs $4.25 per million) and a lower cost per solved task ($0.57 against $0.64).

Model ADeepSeek V4 Pro 0423DeepSeek38.7/1005 independent benchmarks$0.57 per solved taskModel BMuse Spark 1.2Meta55.7/1005 independent benchmarks$0.64 per solved task
Shared results
66
DeepSeek V4 Pro 0423 only
18
Muse Spark 1.2 only
31
Like-for-like categories
8 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 V4 Pro 0423 38.7 from 5 benchmarks, Muse Spark 1.2 55.7 from 5 benchmarks.
Widest category
Agentic work, 18.2 points apart on average over 7 shared benchmarks.
Widest single gap
Design Arena: agents building full-stack apps: DeepSeek V4 Pro 0423 948 against Muse Spark 1.2 1210, 32.7 points apart on a 0 to 100 scale.
Output price
DeepSeek V4 Pro 0423 $2.976 against Muse Spark 1.2 $4.25 per million output tokens, 1.4x apart.
Context window
Both take 1.05M tokens.

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 V4 Pro 0423BarsMuse Spark 1.2Ties
Coding113
Intelligence02
Agentic work07
Reasoning04
Math03
Knowledge010
Web design08
General117
All shared boards264
CodingLike-for-likeSix of these boards feed the Vibe Coding Index113
Coding benchmarks for DeepSeek V4 Pro 0423 and Muse Spark 1.2
BenchmarkDeepSeek V4 Pro 0423Muse Spark 1.2Leader
LiveBench codingNot measured78.2%Only Muse Spark 1.2 measured
LMArena WebDev14471534Muse Spark 1.2
AA Coding Index59.472.2Muse Spark 1.2
LiveCodeBench (Vals AI)87.5%Not measuredOnly DeepSeek V4 Pro 0423 measured
SWE-bench Verified (Vals AI)77.4%86.6%Muse Spark 1.2
Terminal-Bench 4 (Vals AI)11.1%6.1%DeepSeek V4 Pro 0423
Vibe Code Bench (Vals AI)49.9%79.1%Muse Spark 1.2
Code migration (Vals AI)26.2%29.9%Muse Spark 1.2
ProgramBench (Vals AI)0.0%0.5%Muse Spark 1.2
IOI (Vals AI)Not measured21.8%Only Muse Spark 1.2 measured
Terminal-Bench 2.1 (Vals AI)50.2%69.7%Muse Spark 1.2
SkillsBench (Vals AI)51.3%53.0%Muse Spark 1.2
CyberBench (Vals AI)Not measured69.5%Only Muse Spark 1.2 measured
SciCodeNot measured57.4%Only Muse Spark 1.2 measured
FrontierCode17.6%Not measuredOnly DeepSeek V4 Pro 0423 measured
FrontierSWENot measured12.0%Only Muse Spark 1.2 measured
WeirdML48.9%60.3%Muse Spark 1.2
ALE-Bench1006.1Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: coding14701505Muse Spark 1.2
LMArena Text: software and IT14691505Muse Spark 1.2
LMArena WebDev: React14411526Muse Spark 1.2
LMArena WebDev: HTML14351543Muse 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 DeepSeek V4 Pro 0423 and Muse Spark 1.2
BenchmarkDeepSeek V4 Pro 0423Muse Spark 1.2Leader
AA Intelligence Index30.439.6Muse Spark 1.2
Epoch Capabilities Index149.1155.1Muse Spark 1.2
Agentic workLike-for-likePublic boards, shown as published, kept out of the index07
Agentic work benchmarks for DeepSeek V4 Pro 0423 and Muse Spark 1.2
BenchmarkDeepSeek V4 Pro 0423Muse Spark 1.2Leader
AA Agentic Index26.343.2Muse Spark 1.2
Design Arena: agents building web apps10001221Muse Spark 1.2
Design Arena: agents building full-stack apps9481210Muse Spark 1.2
Design Arena: agents building mobile appsNot measured1181Only Muse Spark 1.2 measured
Design Arena: agents building Android appsNot measured1222Only Muse Spark 1.2 measured
Design Arena: agents building Godot games1059Not measuredOnly DeepSeek V4 Pro 0423 measured
Design Arena: agents building gamesNot measured1179Only Muse Spark 1.2 measured
Design Arena: HTML slidesNot measured1085Only Muse Spark 1.2 measured
Design Arena: PowerPoint slidesNot measured1172Only Muse Spark 1.2 measured
LiveBench researchNot measured67.7%Only Muse Spark 1.2 measured
LiveBench browserNot measured36.0%Only Muse Spark 1.2 measured
LiveBench long agenticNot measured21.3%Only Muse Spark 1.2 measured
SREBench (Vals AI)Not measured0.0%Only Muse Spark 1.2 measured
Finance Agent (Vals AI)44.1%60.6%Muse Spark 1.2
Tax Agent Bench (Vals AI)25.2%32.8%Muse Spark 1.2
Legal Research Bench (Vals AI)23.1%43.8%Muse Spark 1.2
Excel Modeling Benchmark (Vals AI)51.6%57.0%Muse Spark 1.2
BioMysteryBench (Vals AI)Not measured64.8%Only Muse Spark 1.2 measured
APEX-AgentsNot measured45.4%Only Muse Spark 1.2 measured
Vending-Bench 2$3,285Not measuredOnly DeepSeek V4 Pro 0423 measured
Neither model has a result on the other 21 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index04
Reasoning benchmarks for DeepSeek V4 Pro 0423 and Muse Spark 1.2
BenchmarkDeepSeek V4 Pro 0423Muse Spark 1.2Leader
GPQA Diamond89.4%Not measuredOnly DeepSeek V4 Pro 0423 measured
LiveBench reasoningNot measured90.0%Only Muse Spark 1.2 measured
MMMU Pro (Vals AI)Not measured86.1%Only Muse Spark 1.2 measured
GPQA Diamond (Artificial Analysis run)Not measured90.4%Only Muse Spark 1.2 measured
Humanity's Last Exam (Artificial Analysis run)Not measured45.5%Only Muse Spark 1.2 measured
GPQA Diamond (Epoch AI run)90.9%Not measuredOnly DeepSeek V4 Pro 0423 measured
SimpleBenchNot measured74.5%Only Muse Spark 1.2 measured
Chess puzzles19.0%27.0%Muse Spark 1.2
Mystery game puzzles17.0%Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: hard prompts14611489Muse Spark 1.2
LMArena Text: expert14631475Muse Spark 1.2
LMArena Text: hard prompts in English14661487Muse Spark 1.2
LMArena Vision: diagramsNot measured1317Only Muse Spark 1.2 measured
Kagi LLM Benchmark53.5%Not measuredOnly DeepSeek V4 Pro 0423 measured
Neither model has a result on the other 8 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index03
Math benchmarks for DeepSeek V4 Pro 0423 and Muse Spark 1.2
BenchmarkDeepSeek V4 Pro 0423Muse Spark 1.2Leader
OTIS Mock AIME96.7%Not measuredOnly DeepSeek V4 Pro 0423 measured
FrontierMath, tiers 1 to 345.3%Not measuredOnly DeepSeek V4 Pro 0423 measured
FrontierMath, tier 42.4%Not measuredOnly DeepSeek V4 Pro 0423 measured
ProofBench16.0%43.0%Muse Spark 1.2
LMArena Text: math14431484Muse Spark 1.2
LMArena Text: mathematical work14551480Muse Spark 1.2
Neither model has a result on the other 7 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index010
Knowledge benchmarks for DeepSeek V4 Pro 0423 and Muse Spark 1.2
BenchmarkDeepSeek V4 Pro 0423Muse Spark 1.2Leader
MMLU Pro87.3%88.3%Muse Spark 1.2
LegalBench (Vals AI)80.3%85.3%Muse Spark 1.2
CaseLaw (Vals AI)59.4%Not measuredOnly DeepSeek V4 Pro 0423 measured
CorpFin (Vals AI)61.4%70.9%Muse Spark 1.2
TaxEval (Vals AI)72.1%80.4%Muse Spark 1.2
MortgageTax (Vals AI)Not measured65.4%Only Muse Spark 1.2 measured
MedCode (Vals AI)40.5%49.4%Muse Spark 1.2
SimpleQA Verified47.0%60.3%Muse Spark 1.2
LMArena Text: science14701500Muse Spark 1.2
LMArena Text: medicine and healthcare14661496Muse Spark 1.2
LMArena Text: legal and government14641517Muse Spark 1.2
LMArena Text: business and finance14431484Muse Spark 1.2
Neither model has a result on the other 6 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index08
Web design benchmarks for DeepSeek V4 Pro 0423 and Muse Spark 1.2
BenchmarkDeepSeek V4 Pro 0423Muse Spark 1.2Leader
Design Arena: websites12341312Muse Spark 1.2
Design Arena: UI components12191298Muse Spark 1.2
Design Arena: game dev12281301Muse Spark 1.2
Design Arena: data viz12031334Muse Spark 1.2
Design Arena: 3D12531301Muse Spark 1.2
Design Arena: all code categories12401312Muse Spark 1.2
Design Arena: SVG11391284Muse Spark 1.2
Design Arena: ASCII art11491284Muse Spark 1.2
LiveBench designNot measured67.1%Only Muse Spark 1.2 measured
GeneralLike-for-likePublic boards, shown as published, kept out of the index117
General benchmarks for DeepSeek V4 Pro 0423 and Muse Spark 1.2
BenchmarkDeepSeek V4 Pro 0423Muse Spark 1.2Leader
LiveBench data analysisNot measured76.5%Only Muse Spark 1.2 measured
Public Benefits Bench (Vals AI)62.9%68.5%Muse Spark 1.2
SAGE (Vals AI)Not measured47.7%Only Muse Spark 1.2 measured
MedScribe (Vals AI)75.1%90.1%Muse Spark 1.2
Vals Multimodal IndexNot measured69.8%Only Muse Spark 1.2 measured
Long context reasoning (AA-LCR)Not measured79.0%Only Muse Spark 1.2 measured
DTBench90.7%94.7%Muse Spark 1.2
LMCA41.2%48.4%Muse Spark 1.2
Surface Evolver Bench40.0%Not measuredOnly DeepSeek V4 Pro 0423 measured
CL-bench Life13.5%Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text14511482Muse Spark 1.2
LMArena Text: instruction following14451462Muse Spark 1.2
LMArena Text: creative writing14461436DeepSeek V4 Pro 0423
LMArena VisionNot measured1304Only Muse Spark 1.2 measured
LMArena Text: English14591483Muse Spark 1.2
LMArena Text: non-English14391476Muse Spark 1.2
LMArena Text: Chinese14811509Muse Spark 1.2
LMArena Text: German1458Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: Japanese1446Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: Korean14091462Muse Spark 1.2
LMArena Text: French14731495Muse Spark 1.2
LMArena Text: Spanish14571484Muse Spark 1.2
LMArena Text: Russian14531483Muse Spark 1.2
LMArena Text: Polish1461Not measuredOnly DeepSeek V4 Pro 0423 measured
LMArena Text: multi-turn14681502Muse Spark 1.2
LMArena Text: longer queries14581473Muse Spark 1.2
LMArena Text: writing and language14471453Muse Spark 1.2
LMArena Text: entertainment and media14311443Muse Spark 1.2
LMArena Vision: EnglishNot measured1291Only Muse Spark 1.2 measured
LMArena Vision: reading textNot measured1313Only Muse Spark 1.2 measured
Neither model has a result on the other 16 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for DeepSeek V4 Pro 0423 and Muse Spark 1.2
BenchmarkDeepSeek V4 Pro 0423Muse 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

DeepSeek V4 Pro 0423$0.0028Fits in one request
Muse Spark 1.2$0.0034Fits in one request

DeepSeek V4 Pro 0423 costs less

Repository review

50K input + 3K output tokens

DeepSeek V4 Pro 0423$0.077Fits in one request
Muse Spark 1.2$0.075Fits in one request

Muse Spark 1.2 costs less

Agent loop

200K input + 10K output tokens

DeepSeek V4 Pro 0423$0.301Fits in one request
Muse Spark 1.2$0.292Fits in one request

Muse Spark 1.2 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 V4 Pro 0423 and Muse Spark 1.2
SpecificationDeepSeek V4 Pro 0423Meta Muse Spark 1.2
MakerDeepSeekMeta
WeightsOpen weightsNo public weights listed
ReleasedApril 24, 2026August 5, 2026
Context window1.05M1.05M
Input $/Mtok$1.358$1.25
Output $/Mtok$2.976$4.25
Coding index38.755.7
Independent measurements behind it55
Cost per solved task$0.57$0.64
SourceOpenRouter listingOpenRouter listing

Questions

Is DeepSeek V4 Pro 0423 better than Muse Spark 1.2 for coding?

Pick Muse Spark 1.2 for coding: it scores 55.7 on the Vibe Coding Index against 38.7 for DeepSeek V4 Pro 0423. Pick DeepSeek V4 Pro 0423 for output tokens 1.4x cheaper ($2.976 vs $4.25 per million) and a lower cost per solved task ($0.57 against $0.64).

Which is cheaper, DeepSeek V4 Pro 0423 or Muse Spark 1.2?

DeepSeek V4 Pro 0423 costs $1.358 per million input tokens and $2.976 per million output tokens; Muse Spark 1.2 costs $1.25 and $4.25. Per solved SWE-bench Verified task, DeepSeek V4 Pro 0423 costs $0.57 and Muse Spark 1.2 $0.64.

Which has the bigger context window, DeepSeek V4 Pro 0423 or Muse Spark 1.2?

Neither: both take 1.05M tokens.

How many benchmarks is the DeepSeek V4 Pro 0423 vs Muse Spark 1.2 comparison based on?

66 benchmarks have a result for both models as of October 22, 2026. DeepSeek V4 Pro 0423's coding index rests on 5 independent benchmarks and Muse Spark 1.2's on 5. 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-v4-pro-vs-meta-muse-spark-1-2" title="DeepSeek V4 Pro 0423 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/deepseek-v4-pro-vs-meta-muse-spark-1-2">LLMBenchmark.ai, DeepSeek V4 Pro 0423 vs Muse Spark 1.2</a></p>