Muse Spark 1.2 vs Hy4 Preview: Benchmark Comparison

Muse Spark 1.2 has the higher score on 7 of 14 shared benchmarks; Hy4 Preview leads on 7.

The largest observed score gap is 37.55 pts on IOI , where Hy4 Preview leads.

Reported ±1 standard-error ranges overlap on 7 of 14 shared benchmarks with comparable uncertainty data. This is not a pairwise statistical significance test.

Shared benchmark results

Scores and reported standard errors come from the latest visible Vals benchmark version available for both models.

Benchmark Muse Spark 1.2 Hy4 Preview Gap Reported uncertainty
Legal Research Bench 43.75% ±3.45 45.19% ±3.46 1.44 pts Reported ±1 SE ranges overlap
LegalBench 85.26% ±0.45 83.76% ±0.41 1.50 pts Reported ±1 SE ranges do not overlap
Tax Agent Bench 32.80% ±2.75 32.62% ±3.38 0.18 pts Reported ±1 SE ranges overlap
MedCode 49.35% ±2.19 43.25% ±2.13 6.10 pts Reported ±1 SE ranges do not overlap
MedScribe 90.06% ±1.96 83.60% ±2.06 6.46 pts Reported ±1 SE ranges do not overlap
ProofBench v1.1 43.00% ±4.98 75.00% ±4.35 32.00 pts Reported ±1 SE ranges do not overlap
BioMysteryBench 64.81% ±1.33 69.26% ±2.59 4.44 pts Reported ±1 SE ranges do not overlap
IOI 21.78% ±0.87 59.33% ±4.60 37.55 pts Reported ±1 SE ranges do not overlap
ProgramBench 0.50% ±0.50 0.00% ±0.00 0.50 pts Reported ±1 SE ranges overlap
Terminal-Bench 4.0 6.06% ±1.51 8.08% ±1.34 2.02 pts Reported ±1 SE ranges overlap
Vibe Code Bench v1.1 79.10% ±3.31 77.48% ±4.04 1.62 pts Reported ±1 SE ranges overlap
CyberBench v1.1 69.52% ±5.56 65.36% ±5.55 4.17 pts Reported ±1 SE ranges overlap
SRE Bench 0.00% ±0.00 2.29% ±0.93 2.29 pts Reported ±1 SE ranges do not overlap
Public Benefits Bench v1.1 68.47% ±1.21 68.61% ±1.21 0.13 pts Reported ±1 SE ranges overlap

Performance by category

Category Muse Spark 1.2 average Hy4 Preview average
Legal 64.51% 64.47%
Finance 32.80% 32.62%
Healthcare 69.70% 63.42%
Math 43.00% 75.00%
Science 64.81% 69.26%
Coding 26.86% 36.22%
Cyber 34.76% 33.82%
Social Mobility 68.47% 68.61%

Cost and latency

Cost per test appears only where the benchmark reports it for both models. Latency is the measured completion time for that benchmark.

Benchmark Muse Spark 1.2 cost Hy4 Preview cost Muse Spark 1.2 latency Hy4 Preview latency
Legal Research Bench $0.51 $0.85 5m07s 1h04m
LegalBench N/A N/A 30.10s 90.56s
Tax Agent Bench $0.27 $0.59 119.91s 41m04s
MedCode N/A N/A 60.17s 6m39s
MedScribe N/A N/A 61.46s 5m50s
ProofBench v1.1 $0.44 $0.34 7m49s 22m54s
BioMysteryBench $0.99 $0.36 23m37s 23m30s
IOI $2.64 $1.35 25m35s 1h01m
ProgramBench $2.47 $12.81 26m38s 2h10m
Terminal-Bench 4.0 $4.14 $3.07 1h18m 2h10m
Vibe Code Bench v1.1 $1.53 $2.18 19m55s 41m33s
CyberBench v1.1 $2.15 $0.46 20m21s 36m43s
SRE Bench $23.43 $4.90 1h40m 1h40m
Public Benefits Bench v1.1 $0.59 $0.25 8m55s 1h03m

Results available only for Muse Spark 1.2

  • Immigration
  • MortgageTax
  • TaxEval v2
  • MMLU Pro
  • MMMU Pro
  • SkillsBench
  • SWE-bench
  • Financial Coaching Bench

Results available only for Hy4 Preview

  • Vals Index
  • Vals Finance Index
  • Vals Legal Index
  • HLAB-Vals-1.1
  • EMB v1.1
  • Finance Agent v2.1
  • Terminal-Bench-Science
  • Code Migration v1.1
Model details Muse Spark 1.2 Model details Hy4 Preview