Claude Haiku 5.5 vs Muse Spark 1.3: Benchmark Comparison
Claude Haiku 5.5 has the higher score on 10 of 14 shared benchmarks; Muse Spark 1.3 leads on 4.
The largest observed score gap is 34.00 pts on ProofBench v1.1 , where Claude Haiku 5.5 leads.
Reported ±1 standard-error ranges overlap on 7 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 | Claude Haiku 5.5 | Muse Spark 1.3 | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Index | 54.44% ±1.41 | 53.20% ±1.07 | 1.24 pts | Reported ±1 SE ranges overlap |
| Harvey's Legal Agent Benchmark | 1.25% ±0.72 | 22.92% ±3.42 | 21.67 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 43.27% ±3.44 | 40.87% ±3.42 | 2.40 pts | Reported ±1 SE ranges overlap |
| EMB | 59.02% ±4.17 | 62.71% ±2.96 | 3.70 pts | Reported ±1 SE ranges overlap |
| Finance Agent (v2) | 54.14% ±2.13 | 58.90% ±0.44 | 4.76 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 31.92% ±3.35 | 35.97% ±3.42 | 4.05 pts | Reported ±1 SE ranges overlap |
| ProofBench v1.1 | 89.00% ±3.15 | 55.00% ±5.00 | 34.00 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench-Science | 10.00% ±3.61 | 4.29% ±2.44 | 5.71 pts | Reported ±1 SE ranges overlap |
| Code Migration | 44.57% ±4.23 | 27.58% ±4.07 | 16.99 pts | Reported ±1 SE ranges do not overlap |
| IOI | 47.28% ±10.72 | 43.94% ±1.45 | 3.33 pts | Reported ±1 SE ranges overlap |
| ProgramBench | 2.50% ±1.11 | 0.50% ±0.50 | 2.00 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench 4.0 | 36.36% ±2.31 | 10.61% ±1.75 | 25.76 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 90.44% ±1.73 | 82.86% ±2.90 | 7.59 pts | Reported ±1 SE ranges do not overlap |
| CyberBench v1.1 | 75.48% ±5.21 | 69.41% ±5.76 | 6.07 pts | Reported ±1 SE ranges overlap |
Performance by category
| Category | Claude Haiku 5.5 average | Muse Spark 1.3 average |
|---|---|---|
| Legal | 22.26% | 31.89% |
| Finance | 48.36% | 52.53% |
| Math | 89.00% | 55.00% |
| Science | 10.00% | 4.29% |
| Coding | 44.23% | 33.10% |
| Cyber | 75.48% | 69.41% |
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 | Claude Haiku 5.5 cost | Muse Spark 1.3 cost | Claude Haiku 5.5 latency | Muse Spark 1.3 latency |
|---|---|---|---|---|
| Vals Index | $3.00 | $3.38 | 47m24s | 32m46s |
| Harvey's Legal Agent Benchmark | $3.39 | $2.76 | 35m14s | 17m04s |
| Legal Research Bench | $1.63 | $0.65 | 1h22m | 8m51s |
| EMB | $1.15 | $3.55 | 18m03s | 22m55s |
| Finance Agent (v2) | $1.64 | $0.74 | 30m49s | 5m42s |
| Tax Agent Bench | $1.16 | $0.32 | 49m34s | 8m25s |
| ProofBench v1.1 | $0.17 | $0.43 | 6m58s | 8m54s |
| Terminal-Bench-Science | $4.52 | $4.41 | 1h34m | 1h08m |
| Code Migration | $6.23 | $4.62 | 1h23m | 1h08m |
| IOI | $0.83 | $2.36 | 12m34s | 29m51s |
| ProgramBench | $6.83 | $11.56 | 1h13m | 1h09m |
| Terminal-Bench 4.0 | $3.12 | $7.04 | 51m56s | 1h59m |
| Vibe Code Bench v1.1 | $6.07 | $2.10 | 34m36s | 12m54s |
| CyberBench v1.1 | $1.25 | $4.26 | 32m39s | 24m53s |
Results available only for Claude Haiku 5.5
- Vals Science Index
- MedCode
- MedScribe
- BioMysteryBench
- MysteryMechanism
- SAGE
- SWE-bench Multimodal
- Vibe Code Bench 1-100
- SRE Bench
- Public Benefits Bench v1.1
Results available only for Muse Spark 1.3
None.