DeepSeek V4.1 Flash vs Muse Spark 1.2: Benchmark Comparison
DeepSeek V4.1 Flash has the higher score on 12 of 20 shared benchmarks; Muse Spark 1.2 leads on 7.
The largest observed score gap is 18.75 pts on Harvey's Legal Agent Benchmark , where Muse Spark 1.2 leads.
Reported ±1 standard-error ranges overlap on 7 of 20 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 | DeepSeek V4.1 Flash | Muse Spark 1.2 | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Index | 51.32% ±1.13 | 49.29% ±1.10 | 2.03 pts | Reported ±1 SE ranges overlap |
| Harvey's Legal Agent Benchmark | 6.67% ±1.83 | 25.42% ±3.67 | 18.75 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 41.35% ±3.42 | 43.75% ±3.45 | 2.40 pts | Reported ±1 SE ranges overlap |
| LegalBench | 83.28% ±0.46 | 85.26% ±0.45 | 1.98 pts | Reported ±1 SE ranges do not overlap |
| EMB | 57.21% ±3.23 | 56.98% ±3.13 | 0.23 pts | Reported ±1 SE ranges overlap |
| Finance Agent (v2) | 53.48% ±0.39 | 60.60% ±0.28 | 7.12 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 62.46% ±3.15 | 56.86% ±2.32 | 5.60 pts | Reported ±1 SE ranges do not overlap |
| MedCode | 41.17% ±2.04 | 49.35% ±2.19 | 8.17 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 85.50% ±1.92 | 90.06% ±1.96 | 4.56 pts | Reported ±1 SE ranges do not overlap |
| ProofBench v1.1 | 54.00% ±5.01 | 43.00% ±4.98 | 11.00 pts | Reported ±1 SE ranges do not overlap |
| BioMysteryBench | 67.78% ±1.11 | 64.81% ±1.33 | 2.96 pts | Reported ±1 SE ranges do not overlap |
| SAGE | 47.88% ±3.44 | 47.66% ±3.44 | 0.22 pts | Reported ±1 SE ranges overlap |
| Code Migration | 45.62% ±4.29 | 29.95% ±4.02 | 15.67 pts | Reported ±1 SE ranges do not overlap |
| IOI | 40.28% ±2.56 | 21.78% ±0.87 | 18.50 pts | Reported ±1 SE ranges do not overlap |
| ProgramBench | 0.50% ±0.50 | 0.50% ±0.50 | 0.00 pts | Reported ±1 SE ranges overlap |
| SkillsBench | 69.80% ±3.88 | 53.04% ±4.42 | 16.77 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench 4.0 | 19.70% ±1.75 | 6.06% ±1.51 | 13.64 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 84.74% ±2.89 | 79.10% ±3.31 | 5.64 pts | Reported ±1 SE ranges overlap |
| CyberBench v1.1 | 73.69% ±5.48 | 69.52% ±5.56 | 4.17 pts | Reported ±1 SE ranges overlap |
| Public Benefits Bench v1.1 | 64.28% ±1.25 | 68.47% ±1.21 | 4.20 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | DeepSeek V4.1 Flash average | Muse Spark 1.2 average |
|---|---|---|
| Legal | 43.76% | 51.48% |
| Finance | 57.72% | 58.15% |
| Healthcare | 63.34% | 69.70% |
| Math | 54.00% | 43.00% |
| Science | 67.78% | 64.81% |
| Education | 47.88% | 47.66% |
| Coding | 43.44% | 31.74% |
| Cyber | 73.69% | 69.52% |
| Social Mobility | 64.28% | 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 | DeepSeek V4.1 Flash cost | Muse Spark 1.2 cost | DeepSeek V4.1 Flash latency | Muse Spark 1.2 latency |
|---|---|---|---|---|
| Vals Index | $0.33 | $1.94 | 27m53s | 22m56s |
| Harvey's Legal Agent Benchmark | $0.17 | $2.09 | 9m39s | 24m16s |
| Legal Research Bench | $0.25 | $0.51 | 11m42s | 5m07s |
| LegalBench | N/A | N/A | 6.53s | 30.10s |
| EMB | $0.23 | $2.31 | 13m00s | 17m07s |
| Finance Agent (v2) | $0.21 | $0.77 | 6m04s | 5m04s |
| Tax Agent Bench | $0.13 | $0.27 | 5m49s | 119.91s |
| MedCode | N/A | N/A | 35.13s | 60.17s |
| MedScribe | N/A | N/A | 53.83s | 61.46s |
| ProofBench v1.1 | $0.13 | $0.44 | 9m07s | 7m49s |
| BioMysteryBench | $0.23 | $0.99 | 14m50s | 23m37s |
| SAGE | N/A | N/A | 35.78s | 29.07s |
| Code Migration | $0.94 | $3.78 | 1h45m | 31m07s |
| IOI | $0.24 | $2.64 | 18m30s | 25m35s |
| ProgramBench | $0.90 | $2.47 | 1h02m | 26m38s |
| SkillsBench | $0.09 | $0.65 | 3m43s | 13m00s |
| Terminal-Bench 4.0 | $0.50 | $4.14 | 48m22s | 1h18m |
| Vibe Code Bench v1.1 | $0.41 | $1.53 | 15m28s | 19m55s |
| CyberBench v1.1 | $0.07 | $2.15 | 13m50s | 20m21s |
| Public Benefits Bench v1.1 | $0.07 | $0.59 | 9m59s | 8m55s |
Results available only for DeepSeek V4.1 Flash
- MysteryMechanism
- Terminal-Bench Science
- Vibe Code Bench 1-100
- SRE Bench
Results available only for Muse Spark 1.2
- MortgageTax
- TaxEval v2
- MMLU Pro
- MMMU Pro
- SWE-bench