Muse Spark 1.3 Max vs GPT-6 Sol: Benchmark Comparison
Muse Spark 1.3 Max has the higher score on 7 of 16 shared benchmarks; GPT-6 Sol leads on 9.
The largest observed score gap is 26.44 pts on Legal Research Bench , where Muse Spark 1.3 Max leads.
Reported ±1 standard-error ranges overlap on 6 of 15 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 Max | GPT-6 Sol | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Index | 58.16% ±1.19 | 57.54% ±1.01 | 0.63 pts | Reported ±1 SE ranges overlap |
| Vals RSI Index | 19.64% | 28.11% | 8.47 pts | Uncertainty comparison unavailable |
| Harvey's Legal Agent Benchmark | 23.75% ±3.55 | 1.67% ±0.82 | 22.08 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 55.29% ±3.46 | 28.85% ±3.15 | 26.44 pts | Reported ±1 SE ranges do not overlap |
| EMB | 67.43% ±3.06 | 71.53% ±2.35 | 4.10 pts | Reported ±1 SE ranges overlap |
| Finance Agent (v2) | 59.96% ±2.06 | 49.05% ±0.58 | 10.91 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 72.44% ±2.88 | 53.05% ±3.29 | 19.40 pts | Reported ±1 SE ranges do not overlap |
| ProofBench v1.1 | 58.00% ±4.96 | 83.00% ±3.77 | 25.00 pts | Reported ±1 SE ranges do not overlap |
| MysteryMechanism | 36.04% ±3.23 | 30.18% ±3.09 | 5.86 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench Science | 10.00% ±3.61 | 30.00% ±5.52 | 20.00 pts | Reported ±1 SE ranges do not overlap |
| Code Migration | 47.41% ±4.26 | 57.20% ±4.21 | 9.78 pts | Reported ±1 SE ranges do not overlap |
| IOI | 56.56% ±2.52 | 82.61% ±9.24 | 26.06 pts | Reported ±1 SE ranges do not overlap |
| ProgramBench | 2.50% ±1.11 | 2.00% ±0.99 | 0.50 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench 4.0 | 24.75% ±0.51 | 44.44% ±3.54 | 19.70 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 85.86% ±2.51 | 87.82% ±2.53 | 1.97 pts | Reported ±1 SE ranges overlap |
| CyberBench v1.1 | 72.74% ±5.67 | 77.98% ±5.11 | 5.24 pts | Reported ±1 SE ranges overlap |
Performance by category
| Category | Muse Spark 1.3 Max average | GPT-6 Sol average |
|---|---|---|
| Legal | 39.52% | 15.26% |
| Finance | 66.61% | 57.88% |
| Math | 58.00% | 83.00% |
| Science | 23.02% | 30.09% |
| Coding | 43.41% | 54.81% |
| Cyber | 72.74% | 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 Max cost | GPT-6 Sol cost | Muse Spark 1.3 Max latency | GPT-6 Sol latency |
|---|---|---|---|---|
| Vals Index | $3.79 | $7.58 | 23m33s | 29m11s |
| Vals RSI Index | $48.82 | $665.11 | 90h00m | 90h00m |
| Harvey's Legal Agent Benchmark | $2.26 | $3.36 | 14m35s | 13m40s |
| Legal Research Bench | $0.59 | $4.90 | 5m17s | 24m09s |
| EMB | $2.55 | $1.28 | 13m50s | 10m22s |
| Finance Agent (v2) | $0.76 | $2.12 | 3m19s | 9m00s |
| Tax Agent Bench | $0.39 | $2.28 | 4m09s | 17m32s |
| ProofBench v1.1 | $0.48 | $0.48 | 9m29s | 5m05s |
| MysteryMechanism | $0.85 | $0.46 | 7m32s | 6m13s |
| Terminal-Bench Science | $6.09 | $5.82 | 1h24m | 1h22m |
| Code Migration | $14.21 | $15.68 | 1h24m | 1h26m |
| IOI | $1.73 | $2.70 | 28m58s | 30m38s |
| ProgramBench | $54.29 | $11.67 | 5h28m | 46m11s |
| Terminal-Bench 4.0 | $6.65 | $5.79 | 52m56s | 34m56s |
| Vibe Code Bench v1.1 | $2.54 | $26.36 | 16m18s | 38m56s |
| CyberBench v1.1 | $3.34 | $1.34 | 16m18s | 12m14s |
Results available only for Muse Spark 1.3 Max
- Vibe Code Bench 1-100
- CUA-bench
Results available only for GPT-6 Sol
- MedCode
- MedScribe
- BioMysteryBench
- SAGE
- Public Benefits Bench v1.1