The Model Index
Up to four models

DeepSeek-V4-Pro vs Muse Spark 1.2

On the index (GPT-4 = 100), DeepSeek-V4-Pro scores 163 and Muse Spark 1.2 162. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.

Next to DeepSeek-V4-Pro
Presets
DeepSeek-V4-Pro
163147–178
WeightsOpen
ReleasedAug 2026
Price in / out$0.11 / $4.44
Context1.05M
Core tests taken6 of 9
Muse Spark 1.2
162146–178
WeightsClosed
ReleasedAug 2026
Price in / out$1.25 / $4.25
Context1.05M
Core tests taken3 of 9

DeepSeek-V4-Pro and Muse Spark 1.2 are too close to call (0 points apart, within the uncertainty), ahead on 2 of 3 shared core tests, and costs 1.7× less.

Test by testthird-party results

DeepSeek-V4-ProMuse Spark 1.2
0%25%50%75%100%REASONING & KNOWLEDGELiveBench Data AnalysisLiveBench Data AnalysisLiveBench IFLiveBench IFLiveBench LanguageLiveBench LanguageLiveBench ReasoningLiveBench ReasoningMMLU ProMMLU ProSimpleQA VerifiedSimpleQA VerifiedMATHSLiveBench MathematicsLiveBench MathematicsProofBench v1.1ProofBench v1.1Chess puzzlesChess puzzlesCODINGTerminal-Bench 4.0Terminal-Bench 4.0LiveBench Agentic CodingLiveBench Agentic CodingLiveBench CodingLiveBench CodingCode MigrationCode MigrationCyberBench v1.1CyberBench v1.1IOIIOIProgramBenchProgramBenchSWE-benchSWE-benchVibe Code Bench v1.1Vibe Code Bench v1.1WeirdMLWeirdMLAGENTSAPEX-AgentsAPEX-AgentsSkillsBenchSkillsBenchSRE BenchSRE BenchDTBenchDTBenchLMCALMCAPROFESSIONAL WORKFinance Agent (v2)Finance Agent (v2)CorpFin v2CorpFin v2EMBEMBLegalBenchLegalBenchLegal Research BenchLegal Research BenchMedCodeMedCodeMedScribeMedScribeTax Agent BenchTax Agent BenchTaxEval v2TaxEval v2
Every result as a table
TestDeepSeek-V4-ProMuse Spark 1.2
Reasoning & knowledge
LiveBench Data Analysis79.2%76.5%
LiveBench IF67.7%74.3%
LiveBench Language82.1%78.6%
LiveBench Reasoning85.8%90.0%
MMLU Pro87.0%88.3%
SimpleQA Verified52.9%60.3%
Maths
LiveBench Mathematics95.1%91.2%
ProofBench v1.150.0%43.0%
Chess puzzles46.0%27.0%
Coding
Terminal-Bench 4.014.1%6.1%
LiveBench Agentic Coding54.9%57.6%
LiveBench Coding77.2%77.5%
Code Migration41.5%29.9%
CyberBench v1.172.9%69.5%
IOI51.6%21.8%
ProgramBench0.0%0.5%
SWE-bench96.4%86.6%
Vibe Code Bench v1.182.3%79.1%
WeirdML66.2%60.4%
Agents
APEX-Agents47.3%45.4%
SkillsBench53.8%53.0%
SRE Bench1.9%0.0%
DTBench93.9%94.7%
LMCA45.5%48.4%
Professional work
Finance Agent (v2)50.4%60.6%
CorpFin v265.4%70.9%
EMB52.8%57.0%
LegalBench82.4%85.3%
Legal Research Bench40.9%43.8%
MedCode42.5%49.4%
MedScribe80.2%90.1%
Tax Agent Bench26.8%32.8%
TaxEval v273.1%80.4%

Tests at least two of these models have taken. Bold rows are the core tests.

This page as a chartDownload PNG