Gemini 3.7 Flash vs Step 5 Preview: Benchmark Comparison
Gemini 3.7 Flash has the higher score on 4 of 8 shared benchmarks; Step 5 Preview leads on 4.
The largest observed score gap is 29.78 pts on IOI , where Gemini 3.7 Flash leads.
Reported ±1 standard-error ranges overlap on 4 of 8 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 3.7 Flash | Step 5 Preview | Gap | Reported uncertainty |
|---|---|---|---|---|
| Legal Research Bench | 34.62% ±3.31 | 40.38% ±3.41 | 5.77 pts | Reported ±1 SE ranges overlap |
| Tax Agent Bench | 22.80% ±3.03 | 27.27% ±3.11 | 4.48 pts | Reported ±1 SE ranges overlap |
| ProofBench v1.1 | 58.00% ±4.96 | 42.00% ±4.96 | 16.00 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench-Science | 7.14% ±3.10 | 2.86% ±2.01 | 4.29 pts | Reported ±1 SE ranges overlap |
| IOI | 67.83% ±3.61 | 38.06% ±6.69 | 29.78 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench 4.0 | 12.12% ±3.03 | 31.82% ±2.62 | 19.70 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 70.39% ±4.84 | 70.40% ±4.27 | 0.00 pts | Reported ±1 SE ranges overlap |
| SRE Bench | 4.58% ±1.29 | 0.38% ±0.38 | 4.20 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Gemini 3.7 Flash average | Step 5 Preview average |
|---|---|---|
| Legal | 34.62% | 40.38% |
| Finance | 22.80% | 27.27% |
| Math | 58.00% | 42.00% |
| Science | 7.14% | 2.86% |
| Coding | 50.12% | 46.76% |
| Cyber | 4.58% | 0.38% |
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 3.7 Flash cost | Step 5 Preview cost | Gemini 3.7 Flash latency | Step 5 Preview latency |
|---|---|---|---|---|
| Legal Research Bench | $0.76 | $0.84 | 4m49s | 36m08s |
| Tax Agent Bench | $0.48 | $0.55 | 87.24s | 21m29s |
| ProofBench v1.1 | $0.56 | $0.71 | 7m04s | 38m34s |
| Terminal-Bench-Science | $7.19 | $4.25 | 1h20m | 2h30m |
| IOI | $3.40 | $2.03 | 21m09s | 1h20m |
| Terminal-Bench 4.0 | $8.71 | $3.11 | 1h24m | 1h58m |
| Vibe Code Bench v1.1 | $4.83 | $5.86 | 23m43s | 1h31m |
| SRE Bench | $21.47 | $3.07 | 1h50m | 1h34m |
Results available only for Gemini 3.7 Flash
- Vals RSI Index
- LegalBench
- MortgageTax
- TaxEval v2
- MedCode
- MedScribe
- GPQA Diamond
- MMLU Pro
- MMMU Pro
- SAGE
- LiveCodeBench
- ProgramBench
- SkillsBench
- SWE-bench
- CyberBench v1.1
Results available only for Step 5 Preview
- Vals Finance Index
- Vals Science Index
- EMB v1.1
- Finance Agent v2.1
- BioMysteryBench
- MysteryMechanism
- Code Migration v1.1
- SWE-bench Multimodal
- Public Benefits Bench v1.1