Muse Spark 1.3 vs GPT-6 Sol: Benchmark Comparison
Muse Spark 1.3 has the higher score on 4 of 14 shared benchmarks; GPT-6 Sol leads on 10.
The largest observed score gap is 38.67 pts on IOI , where GPT-6 Sol leads.
Reported ±1 standard-error ranges overlap on 2 of 14 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 | Muse Spark 1.3 | GPT-6 Sol | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Index | 53.20% ±1.07 | 57.54% ±1.01 | 4.34 pts | Reported ±1 SE ranges do not overlap |
| Harvey's Legal Agent Benchmark | 22.92% ±3.42 | 1.67% ±0.82 | 21.25 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 40.87% ±3.42 | 28.85% ±3.15 | 12.02 pts | Reported ±1 SE ranges do not overlap |
| EMB | 62.71% ±2.96 | 71.53% ±2.35 | 8.82 pts | Reported ±1 SE ranges do not overlap |
| Finance Agent (v2) | 58.90% ±0.44 | 49.05% ±0.58 | 9.85 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 71.93% ±2.92 | 53.05% ±3.29 | 18.89 pts | Reported ±1 SE ranges do not overlap |
| ProofBench v1.1 | 55.00% ±5.00 | 83.00% ±3.77 | 28.00 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench Science | 4.29% ±2.44 | 30.00% ±5.52 | 25.71 pts | Reported ±1 SE ranges do not overlap |
| Code Migration | 27.58% ±4.07 | 57.20% ±4.21 | 29.62 pts | Reported ±1 SE ranges do not overlap |
| IOI | 43.94% ±1.45 | 82.61% ±9.24 | 38.67 pts | Reported ±1 SE ranges do not overlap |
| ProgramBench | 0.50% ±0.50 | 2.00% ±0.99 | 1.50 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench 4.0 | 10.61% ±1.75 | 44.44% ±3.54 | 33.84 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 82.86% ±2.90 | 87.82% ±2.53 | 4.97 pts | Reported ±1 SE ranges overlap |
| CyberBench v1.1 | 69.41% ±5.76 | 77.98% ±5.11 | 8.57 pts | Reported ±1 SE ranges overlap |
Performance by category
| Category | Muse Spark 1.3 average | GPT-6 Sol average |
|---|---|---|
| Legal | 31.89% | 15.26% |
| Finance | 64.52% | 57.88% |
| Math | 55.00% | 83.00% |
| Science | 4.29% | 30.00% |
| Coding | 33.10% | 54.81% |
| Cyber | 69.41% | 77.98% |
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 | Muse Spark 1.3 cost | GPT-6 Sol cost | Muse Spark 1.3 latency | GPT-6 Sol latency |
|---|---|---|---|---|
| Vals Index | $3.38 | $7.58 | 32m42s | 29m11s |
| Harvey's Legal Agent Benchmark | $2.76 | $3.36 | 16m39s | 13m40s |
| Legal Research Bench | $0.65 | $4.90 | 8m51s | 24m09s |
| EMB | $3.55 | $1.28 | 22m55s | 10m22s |
| Finance Agent (v2) | $0.74 | $2.12 | 5m42s | 9m00s |
| Tax Agent Bench | $0.32 | $2.28 | 8m25s | 17m32s |
| ProofBench v1.1 | $0.43 | $0.48 | 8m54s | 5m05s |
| Terminal-Bench Science | $4.41 | $5.82 | 1h08m | 1h22m |
| Code Migration | $4.62 | $15.68 | 1h08m | 1h26m |
| IOI | $2.36 | $2.70 | 29m51s | 30m38s |
| ProgramBench | $11.56 | $11.67 | 1h09m | 46m11s |
| Terminal-Bench 4.0 | $7.04 | $5.79 | 1h59m | 34m56s |
| Vibe Code Bench v1.1 | $2.10 | $26.36 | 12m54s | 38m56s |
| CyberBench v1.1 | $4.26 | $1.34 | 24m53s | 12m14s |
Results available only for Muse Spark 1.3
None.
Results available only for GPT-6 Sol
- Vals RSI Index
- MedCode
- MedScribe
- BioMysteryBench
- MysteryMechanism
- SAGE
- Public Benefits Bench v1.1