Qwen 3.8 Max vs Muse Spark 1.2: Benchmark Comparison
Qwen 3.8 Max has the higher score on 7 of 19 shared benchmarks; Muse Spark 1.2 leads on 12.
The largest observed score gap is 47.11 pts on IOI , where Qwen 3.8 Max leads.
Reported ±1 standard-error ranges overlap on 8 of 19 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 | Qwen 3.8 Max | Muse Spark 1.2 | Gap | Reported uncertainty |
|---|---|---|---|---|
| Legal Research Bench | 47.60% ±3.47 | 43.75% ±3.45 | 3.85 pts | Reported ±1 SE ranges overlap |
| LegalBench | 83.61% ±0.42 | 85.26% ±0.45 | 1.65 pts | Reported ±1 SE ranges do not overlap |
| MortgageTax | 63.99% ±0.94 | 65.42% ±0.93 | 1.43 pts | Reported ±1 SE ranges overlap |
| Tax Agent Bench | 32.09% ±3.37 | 32.80% ±2.75 | 0.70 pts | Reported ±1 SE ranges overlap |
| TaxEval v2 | 75.55% ±0.84 | 80.38% ±0.76 | 4.82 pts | Reported ±1 SE ranges do not overlap |
| MedCode | 40.67% ±2.03 | 49.35% ±2.19 | 8.68 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 84.95% ±2.00 | 90.06% ±1.96 | 5.11 pts | Reported ±1 SE ranges do not overlap |
| ProofBench v1.1 | 58.00% ±4.96 | 43.00% ±4.98 | 15.00 pts | Reported ±1 SE ranges do not overlap |
| MMLU Pro | 88.60% ±0.31 | 88.28% ±0.32 | 0.33 pts | Reported ±1 SE ranges overlap |
| MMMU Pro | 88.03% ±0.78 | 86.13% ±0.83 | 1.91 pts | Reported ±1 SE ranges do not overlap |
| SAGE | 51.25% ±3.42 | 47.66% ±3.44 | 3.59 pts | Reported ±1 SE ranges overlap |
| IOI | 68.89% ±5.20 | 21.78% ±0.87 | 47.11 pts | Reported ±1 SE ranges do not overlap |
| ProgramBench | 0.00% ±0.00 | 0.50% ±0.50 | 0.50 pts | Reported ±1 SE ranges overlap |
| SkillsBench | 42.01% ±4.43 | 53.04% ±4.42 | 11.02 pts | Reported ±1 SE ranges do not overlap |
| SWE-bench | 85.60% ±1.57 | 86.60% ±1.52 | 1.00 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench 4.0 | 34.34% ±3.94 | 6.06% ±1.51 | 28.28 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 64.70% ±5.36 | 79.10% ±3.31 | 14.40 pts | Reported ±1 SE ranges do not overlap |
| CyberBench v1.1 | 28.57% ±3.31 | 69.52% ±5.56 | 40.95 pts | Reported ±1 SE ranges do not overlap |
| Public Benefits Bench v1.1 | 67.12% ±1.22 | 68.47% ±1.21 | 1.35 pts | Reported ±1 SE ranges overlap |
Performance by category
| Category | Qwen 3.8 Max average | Muse Spark 1.2 average |
|---|---|---|
| Legal | 65.60% | 64.51% |
| Finance | 57.21% | 59.53% |
| Healthcare | 62.81% | 69.70% |
| Math | 58.00% | 43.00% |
| Academic | 88.32% | 87.20% |
| Education | 51.25% | 47.66% |
| Coding | 49.26% | 41.18% |
| Cyber | 28.57% | 69.52% |
| Social Mobility | 67.12% | 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 | Qwen 3.8 Max cost | Muse Spark 1.2 cost | Qwen 3.8 Max latency | Muse Spark 1.2 latency |
|---|---|---|---|---|
| Legal Research Bench | $2.49 | $0.51 | 1h01m | 5m07s |
| LegalBench | N/A | N/A | 36.37s | 30.10s |
| MortgageTax | N/A | N/A | 47.50s | 44.62s |
| Tax Agent Bench | $1.55 | $0.27 | 40m36s | 119.91s |
| TaxEval v2 | N/A | N/A | 4m26s | 48.31s |
| MedCode | N/A | N/A | 6m18s | 60.17s |
| MedScribe | N/A | N/A | 4m11s | 61.46s |
| ProofBench v1.1 | $1.50 | $0.44 | 31m08s | 7m49s |
| MMLU Pro | N/A | N/A | 63.58s | 44.23s |
| MMMU Pro | N/A | N/A | 43.85s | 73.84s |
| SAGE | N/A | N/A | 2m13s | 29.07s |
| IOI | $9.47 | $2.64 | 1h47m | 25m35s |
| ProgramBench | $12.36 | $2.47 | 5h54m | 26m38s |
| SkillsBench | $0.52 | $0.65 | 15m48s | 13m00s |
| SWE-bench | $1.13 | $0.55 | 41m42s | 8m07s |
| Terminal-Bench 4.0 | $10.63 | $4.14 | 2h28m | 1h18m |
| Vibe Code Bench v1.1 | $8.24 | $1.53 | 2h49m | 19m55s |
| CyberBench v1.1 | $0.78 | $2.15 | 9m40s | 20m21s |
| Public Benefits Bench v1.1 | $0.99 | $0.59 | 40m44s | 8m55s |
Results available only for Qwen 3.8 Max
- Vals RSI Index
- EMB v1.1
- MysteryMechanism
- Terminal-Bench-Science
- GPQA Diamond
- LiveCodeBench
- SWE-bench Multimodal
- Vibe Code Bench 1-100
Results available only for Muse Spark 1.2
- BioMysteryBench
- SRE Bench