Gemini 3.1 Pro Preview (02/26) vs Mercury 2.5: Benchmark Comparison
Gemini 3.1 Pro Preview (02/26) has the higher score on 11 of 11 shared benchmarks; Mercury 2.5 leads on 0.
The largest observed score gap is 49.39 pts on IOI , where Gemini 3.1 Pro Preview (02/26) leads.
Reported ±1 standard-error ranges overlap on 1 of 11 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) | Mercury 2.5 | Gap | Reported uncertainty |
|---|---|---|---|---|
| Legal Research Bench | 20.67% ±2.81 | 4.33% ±1.41 | 16.35 pts | Reported ±1 SE ranges do not overlap |
| LegalBench | 87.40% ±0.33 | 83.06% ±0.48 | 4.34 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 9.29% ±2.09 | 2.49% ±1.02 | 6.81 pts | Reported ±1 SE ranges do not overlap |
| MedCode | 59.06% ±2.00 | 31.33% ±1.95 | 27.74 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 76.11% ±1.92 | 55.09% ±2.10 | 21.02 pts | Reported ±1 SE ranges do not overlap |
| ProofBench v1.1 | 26.00% ±4.41 | 3.00% ±1.71 | 23.00 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench-Science | 1.43% ±1.43 | 0.00% ±0.00 | 1.43 pts | Reported ±1 SE ranges overlap |
| IOI | 51.83% ±2.59 | 2.44% ±0.68 | 49.39 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench 4.0 | 2.52% ±0.51 | 0.00% ±0.00 | 2.52 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 32.03% ±4.34 | 3.68% ±1.41 | 28.35 pts | Reported ±1 SE ranges do not overlap |
| Public Benefits Bench v1.1 | 53.79% ±1.30 | 45.26% ±1.29 | 8.52 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Gemini 3.1 Pro Preview (02/26) average | Mercury 2.5 average |
|---|---|---|
| Legal | 54.04% | 43.69% |
| Finance | 9.29% | 2.49% |
| Healthcare | 67.59% | 43.21% |
| Math | 26.00% | 3.00% |
| Science | 1.43% | 0.00% |
| Coding | 28.80% | 2.04% |
| Social Mobility | 53.79% | 45.26% |
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 | Mercury 2.5 cost | Gemini 3.1 Pro Preview (02/26) latency | Mercury 2.5 latency |
|---|---|---|---|---|
| Legal Research Bench | $1.10 | $0.06 | 4m02s | 47.61s |
| LegalBench | N/A | N/A | 10.06s | 3.91s |
| Tax Agent Bench | $0.30 | $0.02 | 100.96s | 29.67s |
| MedCode | N/A | N/A | 38.52s | 7.93s |
| MedScribe | N/A | N/A | 69.14s | 9.99s |
| ProofBench v1.1 | $0.78 | $0.04 | 14m49s | 3m22s |
| Terminal-Bench-Science | $5.91 | $0.18 | 43m58s | 28m14s |
| IOI | $1.25 | $0.16 | 12m03s | 4m03s |
| Terminal-Bench 4.0 | $4.11 | $0.37 | 18m44s | 15m14s |
| Vibe Code Bench v1.1 | $3.83 | $0.38 | 20m12s | 3m13s |
| Public Benefits Bench v1.1 | $0.38 | $0.07 | 3m06s | 68.33s |
Results available only for Gemini 3.1 Pro Preview (02/26)
- Vals Science Index
- MortgageTax
- TaxEval v2
- BioMysteryBench
- MysteryMechanism
- GPQA Diamond
- MMLU Pro
- MMMU Pro
- SAGE
- LiveCodeBench
- ProgramBench
- SWE-bench
- Vibe Code Bench 1-100
- SRE Bench
Results available only for Mercury 2.5
- SkillsBench
- CyberBench v1.1