Ember-1 vs Grok 4.6: Benchmark Comparison
Ember-1 has the higher score on 4 of 12 shared benchmarks; Grok 4.6 leads on 8.
The largest observed score gap is 10.00 pts on ProofBench v1.1 , where Ember-1 leads.
Reported ±1 standard-error ranges overlap on 7 of 12 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 | Grok 4.6 | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Science Index | 31.76% ±1.66 | 38.09% ±1.64 | 6.34 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 44.71% ±3.46 | 48.08% ±3.47 | 3.36 pts | Reported ±1 SE ranges overlap |
| Tax Agent Bench | 31.08% ±3.34 | 33.14% ±3.40 | 2.06 pts | Reported ±1 SE ranges overlap |
| ProofBench v1.1 | 61.00% ±4.90 | 51.00% ±5.02 | 10.00 pts | Reported ±1 SE ranges do not overlap |
| BioMysteryBench | 67.04% ±3.03 | 72.22% ±0.00 | 5.18 pts | Reported ±1 SE ranges do not overlap |
| MysteryMechanism | 22.52% ±2.81 | 30.63% ±3.10 | 8.11 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench-Science | 5.71% ±2.79 | 11.43% ±3.83 | 5.71 pts | Reported ±1 SE ranges overlap |
| IOI | 46.33% ±9.16 | 47.61% ±2.09 | 1.28 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench 4.0 | 20.71% ±3.64 | 17.17% ±1.34 | 3.54 pts | Reported ±1 SE ranges overlap |
| Vibe Code Bench v1.1 | 83.87% ±3.15 | 76.24% ±3.82 | 7.63 pts | Reported ±1 SE ranges do not overlap |
| CyberBench v1.1 | 69.52% ±5.56 | 67.68% ±5.87 | 1.85 pts | Reported ±1 SE ranges overlap |
| Public Benefits Bench v1.1 | 66.17% ±1.23 | 66.85% ±1.23 | 0.68 pts | Reported ±1 SE ranges overlap |
Performance by category
| Category | Ember-1 average | Grok 4.6 average |
|---|---|---|
| Legal | 44.71% | 48.08% |
| Finance | 31.08% | 33.14% |
| Math | 61.00% | 51.00% |
| Science | 31.76% | 38.09% |
| Coding | 50.30% | 47.01% |
| Cyber | 69.52% | 67.68% |
| Social Mobility | 66.17% | 66.85% |
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 | Grok 4.6 cost | Ember-1 latency | Grok 4.6 latency |
|---|---|---|---|---|
| Vals Science Index | $10.99 | $2.45 | 1h18m | 40m10s |
| Legal Research Bench | $3.67 | $1.53 | 38m34s | 25m04s |
| Tax Agent Bench | $2.97 | $0.98 | 30m46s | 10m58s |
| ProofBench v1.1 | $1.74 | $0.76 | 21m09s | 14m22s |
| BioMysteryBench | $0.91 | $1.23 | 9m26s | 14m18s |
| MysteryMechanism | $1.29 | $1.31 | 20m15s | 27m55s |
| Terminal-Bench-Science | $30.76 | $4.82 | 3h23m | 1h18m |
| IOI | $24.46 | $8.18 | 1h47m | 3h57m |
| Terminal-Bench 4.0 | $19.48 | $5.06 | 1h55m | 29m25s |
| Vibe Code Bench v1.1 | $8.56 | $4.88 | 39m18s | 25m30s |
| CyberBench v1.1 | $2.43 | $4.07 | 22m02s | 36m52s |
| Public Benefits Bench v1.1 | $1.02 | $0.92 | 13m21s | 25m53s |
Results available only for Ember-1
None.
Results available only for Grok 4.6
- Vals RSI Index
- LegalBench
- MortgageTax
- TaxEval v2
- MedCode
- MedScribe
- GPQA Diamond
- MMLU Pro
- SAGE
- LiveCodeBench
- ProgramBench
- SkillsBench
- SWE-bench
- Vibe Code Bench 1-100
- Time Horizon Index: KSP