Kimi K3 vs Muse Spark 1.2: Benchmark Comparison
Kimi K3 has the higher score on 17 of 24 shared benchmarks; Muse Spark 1.2 leads on 7.
The largest observed score gap is 44.00 pts on ProofBench v1.1 , where Kimi K3 leads.
Reported ±1 standard-error ranges overlap on 10 of 24 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 | Kimi K3 | Muse Spark 1.2 | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Index | 50.30% ±0.99 | 49.29% ±1.10 | 1.01 pts | Reported ±1 SE ranges overlap |
| Harvey's Legal Agent Benchmark | 12.92% ±2.68 | 25.42% ±3.67 | 12.50 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 46.15% ±3.46 | 43.75% ±3.45 | 2.40 pts | Reported ±1 SE ranges overlap |
| LegalBench | 86.21% ±0.41 | 85.26% ±0.45 | 0.95 pts | Reported ±1 SE ranges do not overlap |
| EMB | 66.68% ±2.87 | 56.98% ±3.13 | 9.70 pts | Reported ±1 SE ranges do not overlap |
| Finance Agent (v2) | 53.11% ±0.37 | 60.60% ±0.28 | 7.48 pts | Reported ±1 SE ranges do not overlap |
| MortgageTax | 66.34% ±0.91 | 65.42% ±0.93 | 0.91 pts | Reported ±1 SE ranges overlap |
| Tax Agent Bench | 68.67% ±3.08 | 56.86% ±2.32 | 11.81 pts | Reported ±1 SE ranges do not overlap |
| TaxEval v2 | 75.72% ±0.84 | 80.38% ±0.76 | 4.66 pts | Reported ±1 SE ranges do not overlap |
| MedCode | 49.36% ±2.20 | 49.35% ±2.19 | 0.01 pts | Reported ±1 SE ranges overlap |
| MedScribe | 88.05% ±1.98 | 90.06% ±1.96 | 2.01 pts | Reported ±1 SE ranges overlap |
| ProofBench v1.1 | 87.00% ±3.38 | 43.00% ±4.98 | 44.00 pts | Reported ±1 SE ranges do not overlap |
| BioMysteryBench | 72.59% ±1.61 | 64.81% ±1.33 | 7.78 pts | Reported ±1 SE ranges do not overlap |
| MMLU Pro | 87.97% ±0.32 | 88.28% ±0.32 | 0.30 pts | Reported ±1 SE ranges overlap |
| MMMU Pro | 88.15% ±0.78 | 86.13% ±0.83 | 2.02 pts | Reported ±1 SE ranges do not overlap |
| SAGE | 52.78% ±3.41 | 47.66% ±3.44 | 5.12 pts | Reported ±1 SE ranges overlap |
| Code Migration | 16.10% ±4.10 | 29.95% ±4.02 | 13.86 pts | Reported ±1 SE ranges do not overlap |
| IOI | 48.94% ±9.82 | 21.78% ±0.87 | 27.17 pts | Reported ±1 SE ranges do not overlap |
| ProgramBench | 2.00% ±0.99 | 0.50% ±0.50 | 1.50 pts | Reported ±1 SE ranges do not overlap |
| SWE-bench | 93.40% ±1.11 | 86.60% ±1.52 | 6.80 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench 4.0 | 17.17% ±0.51 | 6.06% ±1.51 | 11.11 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 84.97% ±2.75 | 79.10% ±3.31 | 5.87 pts | Reported ±1 SE ranges overlap |
| CyberBench v1.1 | 75.24% ±5.56 | 69.52% ±5.56 | 5.71 pts | Reported ±1 SE ranges overlap |
| Public Benefits Bench v1.1 | 68.20% ±1.21 | 68.47% ±1.21 | 0.27 pts | Reported ±1 SE ranges overlap |
Performance by category
| Category | Kimi K3 average | Muse Spark 1.2 average |
|---|---|---|
| Legal | 48.43% | 51.48% |
| Finance | 66.10% | 64.05% |
| Healthcare | 68.70% | 69.70% |
| Math | 87.00% | 43.00% |
| Science | 72.59% | 64.81% |
| Academic | 88.06% | 87.20% |
| Education | 52.78% | 47.66% |
| Coding | 43.76% | 37.33% |
| Cyber | 75.24% | 69.52% |
| Social Mobility | 68.20% | 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 | Kimi K3 cost | Muse Spark 1.2 cost | Kimi K3 latency | Muse Spark 1.2 latency |
|---|---|---|---|---|
| Vals Index | $6.38 | $1.94 | 1h08m | 22m56s |
| Harvey's Legal Agent Benchmark | $3.83 | $2.09 | 13m58s | 24m16s |
| Legal Research Bench | $3.47 | $0.51 | 12m40s | 5m07s |
| LegalBench | N/A | N/A | 7.11s | 30.10s |
| EMB | $3.07 | $2.31 | 12m21s | 17m07s |
| Finance Agent (v2) | $1.91 | $0.77 | 4m46s | 5m04s |
| MortgageTax | N/A | N/A | 78.94s | 44.62s |
| Tax Agent Bench | $2.79 | $0.27 | 40m15s | 119.91s |
| TaxEval v2 | N/A | N/A | 2m25s | 48.31s |
| MedCode | N/A | N/A | 36.99s | 60.17s |
| MedScribe | N/A | N/A | 38.24s | 61.46s |
| ProofBench v1.1 | $1.65 | $0.44 | 34m31s | 7m49s |
| BioMysteryBench | $1.06 | $0.99 | 4m55s | 23m37s |
| MMLU Pro | N/A | N/A | 38.72s | 44.23s |
| MMMU Pro | N/A | N/A | 106.52s | 73.84s |
| SAGE | N/A | N/A | 47.30s | 29.07s |
| Code Migration | $13.87 | $3.78 | 4h18m | 31m07s |
| IOI | $14.17 | $2.64 | 4h18m | 25m35s |
| ProgramBench | $70.48 | $2.47 | 5h42m | 26m38s |
| SWE-bench | $0.76 | $0.55 | 10m19s | 8m07s |
| Terminal-Bench 4.0 | $12.02 | $4.14 | 3h09m | 1h18m |
| Vibe Code Bench v1.1 | $10.01 | $1.53 | 16m39s | 19m55s |
| CyberBench v1.1 | $2.13 | $2.15 | 30m44s | 20m21s |
| Public Benefits Bench v1.1 | $0.91 | $0.59 | 7m19s | 8m55s |
Results available only for Kimi K3
- Vals RSI Index
- Terminal-Bench Science
- GPQA Diamond
- LiveCodeBench
- Vibe Code Bench 1-100
- Time Horizon Index: KSP
Results available only for Muse Spark 1.2
- SkillsBench