Ember-1 vs Muse Spark 1.2: Benchmark Comparison
Ember-1 has the higher score on 6 of 9 shared benchmarks; Muse Spark 1.2 leads on 2.
The largest observed score gap is 24.55 pts on IOI , where Ember-1 leads.
Reported ±1 standard-error ranges overlap on 6 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 | Ember-1 | Muse Spark 1.2 | Gap | Reported uncertainty |
|---|---|---|---|---|
| Legal Research Bench | 44.71% ±3.46 | 43.75% ±3.45 | 0.96 pts | Reported ±1 SE ranges overlap |
| Tax Agent Bench | 31.08% ±3.34 | 32.80% ±2.75 | 1.72 pts | Reported ±1 SE ranges overlap |
| ProofBench v1.1 | 61.00% ±4.90 | 43.00% ±4.98 | 18.00 pts | Reported ±1 SE ranges do not overlap |
| BioMysteryBench | 67.04% ±3.03 | 64.81% ±1.33 | 2.22 pts | Reported ±1 SE ranges overlap |
| IOI | 46.33% ±9.16 | 21.78% ±0.87 | 24.55 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench 4.0 | 20.71% ±3.64 | 6.06% ±1.51 | 14.65 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 83.87% ±3.15 | 79.10% ±3.31 | 4.77 pts | Reported ±1 SE ranges overlap |
| CyberBench v1.1 | 69.52% ±5.56 | 69.52% ±5.56 | 0.00 pts | Reported ±1 SE ranges overlap |
| Public Benefits Bench v1.1 | 66.17% ±1.23 | 68.47% ±1.21 | 2.30 pts | Reported ±1 SE ranges overlap |
Performance by category
| Category | Ember-1 average | Muse Spark 1.2 average |
|---|---|---|
| Legal | 44.71% | 43.75% |
| Finance | 31.08% | 32.80% |
| Math | 61.00% | 43.00% |
| Science | 67.04% | 64.81% |
| Coding | 50.30% | 35.65% |
| Cyber | 69.52% | 69.52% |
| Social Mobility | 66.17% | 68.47% |
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 | Muse Spark 1.2 cost | Ember-1 latency | Muse Spark 1.2 latency |
|---|---|---|---|---|
| Legal Research Bench | $3.67 | $0.51 | 38m34s | 5m07s |
| Tax Agent Bench | $2.97 | $0.27 | 30m46s | 119.91s |
| ProofBench v1.1 | $1.74 | $0.44 | 21m09s | 7m49s |
| BioMysteryBench | $0.91 | $0.99 | 9m26s | 23m37s |
| IOI | $24.46 | $2.64 | 1h47m | 25m35s |
| Terminal-Bench 4.0 | $19.48 | $4.14 | 1h55m | 1h18m |
| Vibe Code Bench v1.1 | $8.56 | $1.53 | 39m18s | 19m55s |
| CyberBench v1.1 | $2.43 | $2.15 | 22m02s | 20m21s |
| Public Benefits Bench v1.1 | $1.02 | $0.59 | 13m21s | 8m55s |
Results available only for Ember-1
- Vals Science Index
- MysteryMechanism
- Terminal-Bench-Science
Results available only for Muse Spark 1.2
- LegalBench
- MortgageTax
- TaxEval v2
- MedCode
- MedScribe
- MMLU Pro
- MMMU Pro
- SAGE
- ProgramBench
- SkillsBench
- SWE-bench
- SRE Bench