Ember-1 vs Gemini 3.7 Flash: Benchmark Comparison
Ember-1 has the higher score on 6 of 8 shared benchmarks; Gemini 3.7 Flash leads on 2.
The largest observed score gap is 25.77 pts on CyberBench v1.1 , where Ember-1 leads.
Reported ±1 standard-error ranges overlap on 2 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 | Ember-1 | Gemini 3.7 Flash | Gap | Reported uncertainty |
|---|---|---|---|---|
| Legal Research Bench | 44.71% ±3.46 | 34.62% ±3.31 | 10.10 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 31.08% ±3.34 | 22.80% ±3.03 | 8.28 pts | Reported ±1 SE ranges do not overlap |
| ProofBench v1.1 | 61.00% ±4.90 | 58.00% ±4.96 | 3.00 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench-Science | 5.71% ±2.79 | 7.14% ±3.10 | 1.43 pts | Reported ±1 SE ranges overlap |
| IOI | 46.33% ±9.16 | 67.83% ±3.61 | 21.50 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench 4.0 | 20.71% ±3.64 | 12.12% ±3.03 | 8.59 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 83.87% ±3.15 | 70.39% ±4.84 | 13.47 pts | Reported ±1 SE ranges do not overlap |
| CyberBench v1.1 | 69.52% ±5.56 | 43.75% ±2.21 | 25.77 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Ember-1 average | Gemini 3.7 Flash average |
|---|---|---|
| Legal | 44.71% | 34.62% |
| Finance | 31.08% | 22.80% |
| Math | 61.00% | 58.00% |
| Science | 5.71% | 7.14% |
| Coding | 50.30% | 50.12% |
| Cyber | 69.52% | 43.75% |
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 | Ember-1 cost | Gemini 3.7 Flash cost | Ember-1 latency | Gemini 3.7 Flash latency |
|---|---|---|---|---|
| Legal Research Bench | $3.67 | $0.76 | 38m34s | 4m49s |
| Tax Agent Bench | $2.97 | $0.48 | 30m46s | 87.24s |
| ProofBench v1.1 | $1.74 | $0.56 | 21m09s | 7m04s |
| Terminal-Bench-Science | $30.76 | $7.19 | 3h23m | 1h20m |
| IOI | $24.46 | $3.40 | 1h47m | 21m09s |
| Terminal-Bench 4.0 | $19.48 | $8.71 | 1h55m | 1h24m |
| Vibe Code Bench v1.1 | $8.56 | $4.83 | 39m18s | 23m43s |
| CyberBench v1.1 | $2.43 | $2.15 | 22m02s | 7m34s |
Results available only for Ember-1
- Vals Science Index
- BioMysteryBench
- MysteryMechanism
- Public Benefits Bench v1.1
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
- SRE Bench