Claude Sonnet 5 vs Ember-1: Benchmark Comparison
Claude Sonnet 5 has the higher score on 1 of 9 shared benchmarks; Ember-1 leads on 7.
The largest observed score gap is 16.00 pts on ProofBench v1.1 , where Claude Sonnet 5 leads.
Reported ±1 standard-error ranges overlap on 7 of 9 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 | Claude Sonnet 5 | Ember-1 | Gap | Reported uncertainty |
|---|---|---|---|---|
| Legal Research Bench | 41.83% ±3.43 | 44.71% ±3.46 | 2.89 pts | Reported ±1 SE ranges overlap |
| Tax Agent Bench | 29.37% ±3.25 | 31.08% ±3.34 | 1.70 pts | Reported ±1 SE ranges overlap |
| ProofBench v1.1 | 77.00% ±4.23 | 61.00% ±4.90 | 16.00 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench-Science | 5.71% ±2.79 | 5.71% ±2.79 | 0.00 pts | Reported ±1 SE ranges overlap |
| IOI | 45.00% ±2.75 | 46.33% ±9.16 | 1.33 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench 4.0 | 10.61% ±1.51 | 20.71% ±3.64 | 10.10 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 81.33% ±3.05 | 83.87% ±3.15 | 2.54 pts | Reported ±1 SE ranges overlap |
| CyberBench v1.1 | 61.91% ±5.74 | 69.52% ±5.56 | 7.62 pts | Reported ±1 SE ranges overlap |
| Public Benefits Bench v1.1 | 66.03% ±1.23 | 66.17% ±1.23 | 0.14 pts | Reported ±1 SE ranges overlap |
Performance by category
| Category | Claude Sonnet 5 average | Ember-1 average |
|---|---|---|
| Legal | 41.83% | 44.71% |
| Finance | 29.37% | 31.08% |
| Math | 77.00% | 61.00% |
| Science | 5.71% | 5.71% |
| Coding | 45.64% | 50.30% |
| Cyber | 61.91% | 69.52% |
| Social Mobility | 66.03% | 66.17% |
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 | Claude Sonnet 5 cost | Ember-1 cost | Claude Sonnet 5 latency | Ember-1 latency |
|---|---|---|---|---|
| Legal Research Bench | $2.72 | $3.67 | 25m46s | 38m34s |
| Tax Agent Bench | $1.79 | $2.97 | 19m19s | 30m46s |
| ProofBench v1.1 | $1.37 | $1.74 | 14m58s | 21m09s |
| Terminal-Bench-Science | $30.30 | $30.76 | 3h06m | 3h23m |
| IOI | $12.89 | $24.46 | 59m44s | 1h47m |
| Terminal-Bench 4.0 | $26.05 | $19.48 | 1h45m | 1h55m |
| Vibe Code Bench v1.1 | $25.39 | $8.56 | 1h07m | 39m18s |
| CyberBench v1.1 | $1.89 | $2.43 | 16m57s | 22m02s |
| Public Benefits Bench v1.1 | $1.29 | $1.02 | 21m09s | 13m21s |
Results available only for Claude Sonnet 5
- LegalBench
- MortgageTax
- TaxEval v2
- MedCode
- MedScribe
- GPQA Diamond
- MMLU Pro
- MMMU Pro
- SAGE
- LiveCodeBench
- ProgramBench
- SkillsBench
- SWE-bench
- Vibe Code Bench 1-100
- SRE Bench
Results available only for Ember-1
- Vals Science Index
- BioMysteryBench
- MysteryMechanism