Gemini 3.1 Pro Preview (02/26) vs MiniMax-M3: Benchmark Comparison

Gemini 3.1 Pro Preview (02/26) has the higher score on 11 of 18 shared benchmarks; MiniMax-M3 leads on 7.

The largest observed score gap is 15.53 pts on Vibe Code Bench v1.1 , where MiniMax-M3 leads.

Reported ±1 standard-error ranges overlap on 6 of 18 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 Gemini 3.1 Pro Preview (02/26) MiniMax-M3 Gap Reported uncertainty
Legal Research Bench 20.67% ±2.81 29.81% ±3.18 9.14 pts Reported ±1 SE ranges do not overlap
LegalBench 87.40% ±0.33 85.42% ±0.44 1.98 pts Reported ±1 SE ranges do not overlap
MortgageTax 69.40% ±0.91 68.36% ±0.91 1.03 pts Reported ±1 SE ranges overlap
Tax Agent Bench 9.29% ±2.09 20.59% ±2.87 11.29 pts Reported ±1 SE ranges do not overlap
TaxEval v2 72.88% ±0.86 72.73% ±0.86 0.15 pts Reported ±1 SE ranges overlap
MedCode 59.06% ±2.00 46.29% ±2.10 12.77 pts Reported ±1 SE ranges do not overlap
MedScribe 76.11% ±1.92 87.25% ±1.96 11.14 pts Reported ±1 SE ranges do not overlap
ProofBench v1.1 26.00% ±4.41 18.00% ±3.86 8.00 pts Reported ±1 SE ranges overlap
GPQA Diamond 95.45% ±1.05 92.68% ±1.44 2.78 pts Reported ±1 SE ranges do not overlap
MMLU Pro 90.99% ±0.28 84.22% ±0.36 6.77 pts Reported ±1 SE ranges do not overlap
MMMU Pro 88.21% ±0.78 81.16% ±0.94 7.05 pts Reported ±1 SE ranges do not overlap
SAGE 83.18% ±1.95 85.53% ±1.75 2.36 pts Reported ±1 SE ranges overlap
LiveCodeBench 88.48% ±0.93 82.15% ±1.05 6.33 pts Reported ±1 SE ranges do not overlap
SWE-bench 78.80% ±1.83 75.00% ±1.94 3.80 pts Reported ±1 SE ranges do not overlap
Terminal-Bench 4.0 2.52% ±0.51 1.01% ±1.01 1.51 pts Reported ±1 SE ranges overlap
Vibe Code Bench v1.1 32.03% ±4.34 47.57% ±5.44 15.53 pts Reported ±1 SE ranges do not overlap
Financial Coaching Bench 66.73% ±1.50 69.05% ±1.58 2.32 pts Reported ±1 SE ranges overlap
Public Benefits Bench v1.1 53.79% ±1.30 64.14% ±1.25 10.35 pts Reported ±1 SE ranges do not overlap

Performance by category

Category Gemini 3.1 Pro Preview (02/26) average MiniMax-M3 average
Legal 54.04% 57.61%
Finance 50.52% 53.89%
Healthcare 67.59% 66.77%
Math 26.00% 18.00%
Academic 91.55% 86.02%
Education 83.18% 85.53%
Coding 50.46% 51.43%
Social Mobility 60.26% 66.59%

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 Gemini 3.1 Pro Preview (02/26) cost MiniMax-M3 cost Gemini 3.1 Pro Preview (02/26) latency MiniMax-M3 latency
Legal Research Bench $1.10 $0.34 4m02s 13m34s
LegalBench N/A N/A 10.06s 8.21s
MortgageTax N/A N/A 23.62s 26.76s
Tax Agent Bench $0.30 $0.16 100.96s 5m15s
TaxEval v2 N/A N/A 41.93s 94.04s
MedCode N/A N/A 38.52s 63.12s
MedScribe N/A N/A 69.14s 2m04s
ProofBench v1.1 $0.78 $0.42 14m49s 11m59s
GPQA Diamond N/A N/A 65.76s 4m39s
MMLU Pro N/A N/A 23.91s 41.39s
MMMU Pro N/A N/A 76.99s 68.35s
SAGE $0.05 $0.01 45.64s 51.87s
LiveCodeBench N/A N/A 88.89s 6m07s
SWE-bench $0.78 $0.42 5m12s 12m06s
Terminal-Bench 4.0 $4.11 $7.23 18m44s 1h16m
Vibe Code Bench v1.1 $3.83 $6.45 20m12s 1h25m
Financial Coaching Bench $0.12 $0.03 7m23s 2m49s
Public Benefits Bench v1.1 $0.38 $0.25 3m06s 11m31s

Results available only for Gemini 3.1 Pro Preview (02/26)

  • Vals Science Index
  • BioMysteryBench
  • MysteryMechanism
  • Terminal-Bench-Science
  • IOI
  • ProgramBench
  • SRE Bench

Results available only for MiniMax-M3

  • Vals Finance Index
  • Vals Inference Provider Index
  • Immigration
  • EMB v1.1
  • Finance Agent v2.1
  • Code Migration v1.1
  • SkillsBench
  • SWE-bench Multimodal
  • CyberBench v1.1
  • Access to Justice Benchmark
Model details Gemini 3.1 Pro Preview (02/26) Model details MiniMax-M3