Mercury 2.5 vs GPT 5.4 Mini: Benchmark Comparison

Mercury 2.5 has the higher score on 0 of 4 shared benchmarks; GPT 5.4 Mini leads on 4.

The largest observed score gap is 44.29 pts on Vibe Code Bench v1.1 , where GPT 5.4 Mini leads.

Reported ±1 standard-error ranges overlap on 0 of 4 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 Mercury 2.5 GPT 5.4 Mini Gap Reported uncertainty
Legal Research Bench 4.33% ±1.41 12.50% ±2.30 8.17 pts Reported ±1 SE ranges do not overlap
Tax Agent Bench 2.49% ±1.02 8.81% ±2.04 6.32 pts Reported ±1 SE ranges do not overlap
Terminal-Bench 4.0 0.00% ±0.00 2.52% ±1.01 2.52 pts Reported ±1 SE ranges do not overlap
Vibe Code Bench v1.1 3.68% ±1.41 47.97% ±5.61 44.29 pts Reported ±1 SE ranges do not overlap

Performance by category

Category Mercury 2.5 average GPT 5.4 Mini average
Legal 4.33% 12.50%
Finance 2.49% 8.81%
Coding 1.84% 25.25%

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 Mercury 2.5 cost GPT 5.4 Mini cost Mercury 2.5 latency GPT 5.4 Mini latency
Legal Research Bench $0.06 $1.86 47.61s 37m36s
Tax Agent Bench $0.02 $0.81 29.67s 16m46s
Terminal-Bench 4.0 $0.37 $1.44 15m14s 29m30s
Vibe Code Bench v1.1 $0.38 $1.19 3m13s 34m15s

Results available only for Mercury 2.5

  • LegalBench
  • MedCode
  • MedScribe
  • ProofBench v1.1
  • Terminal-Bench-Science
  • IOI
  • SkillsBench
  • CyberBench v1.1
  • Public Benefits Bench v1.1

Results available only for GPT 5.4 Mini

  • MortgageTax
  • TaxEval v2
  • GPQA Diamond
  • MMLU Pro
  • MMMU Pro
  • SAGE
  • LiveCodeBench
  • ProgramBench
  • SWE-bench
Model details Mercury 2.5 Model details GPT 5.4 Mini