Claude Haiku 5.5 vs Grok 4.7: Benchmark Comparison
Claude Haiku 5.5 has the higher score on 15 of 25 shared benchmarks; Grok 4.7 leads on 9.
The largest observed score gap is 63.00 pts on ProofBench v1.1 , where Claude Haiku 5.5 leads.
Reported ±1 standard-error ranges overlap on 16 of 25 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 | Grok 4.7 | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Finance Index | 57.45% ±1.80 | 58.46% ±1.45 | 1.01 pts | Reported ±1 SE ranges overlap |
| Vals Index | 58.71% ±1.28 | 58.56% ±1.01 | 0.14 pts | Reported ±1 SE ranges overlap |
| Vals Legal Index | 70.27% ±1.83 | 72.60% ±1.82 | 2.33 pts | Reported ±1 SE ranges overlap |
| Vals Science Index | 38.50% ±1.67 | 34.83% ±1.61 | 3.67 pts | Reported ±1 SE ranges do not overlap |
| HLAB-Vals-1.1 | 4.17% ±1.52 | 18.33% ±3.18 | 14.17 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 43.27% ±3.44 | 47.12% ±3.47 | 3.85 pts | Reported ±1 SE ranges overlap |
| EMB v1.1 | 65.48% ±3.69 | 71.50% ±2.33 | 6.02 pts | Reported ±1 SE ranges do not overlap |
| Finance Agent v2.1 | 65.37% ±2.06 | 60.59% ±2.11 | 4.79 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 31.92% ±3.35 | 34.20% ±3.42 | 2.28 pts | Reported ±1 SE ranges overlap |
| MedCode | 51.54% ±2.17 | 49.55% ±2.17 | 1.99 pts | Reported ±1 SE ranges overlap |
| MedScribe | 89.54% ±1.92 | 89.38% ±1.89 | 0.17 pts | Reported ±1 SE ranges overlap |
| ProofBench v1.1 | 89.00% ±3.15 | 26.00% ±4.41 | 63.00 pts | Reported ±1 SE ranges do not overlap |
| BioMysteryBench | 70.37% ±1.33 | 69.26% ±1.33 | 1.11 pts | Reported ±1 SE ranges overlap |
| MysteryMechanism | 35.13% ±3.21 | 25.23% ±2.92 | 9.91 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench-Science | 10.00% ±3.61 | 10.00% ±3.61 | 0.00 pts | Reported ±1 SE ranges overlap |
| SAGE | 45.78% ±3.42 | 40.79% ±3.35 | 4.99 pts | Reported ±1 SE ranges overlap |
| Code Migration v1.1 | 52.36% ±3.49 | 53.22% ±2.56 | 0.86 pts | Reported ±1 SE ranges overlap |
| IOI | 47.28% ±10.72 | 57.72% ±1.99 | 10.44 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 |
| SWE-bench Multimodal | 43.96% ±2.27 | 42.92% ±2.26 | 1.04 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench 4.0 | 36.36% ±2.31 | 28.79% ±2.31 | 7.58 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 90.44% ±1.73 | 86.17% ±2.18 | 4.27 pts | Reported ±1 SE ranges do not overlap |
| CyberBench v1.1 | 75.48% ±5.21 | 69.46% ±5.67 | 6.01 pts | Reported ±1 SE ranges overlap |
| SRE Bench | 5.72% ±1.44 | 3.82% ±1.19 | 1.91 pts | Reported ±1 SE ranges overlap |
| Public Benefits Bench v1.1 | 63.19% ±1.25 | 65.63% ±1.24 | 2.44 pts | Reported ±1 SE ranges overlap |
Performance by category
| Category | Claude Haiku 5.5 average | Grok 4.7 average |
|---|---|---|
| Legal | 23.72% | 32.72% |
| Finance | 54.26% | 55.43% |
| Healthcare | 70.54% | 69.47% |
| Math | 89.00% | 26.00% |
| Science | 38.50% | 34.83% |
| Education | 45.78% | 40.79% |
| Coding | 45.48% | 44.89% |
| Cyber | 40.60% | 36.64% |
| Social Mobility | 63.19% | 65.63% |
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 | Grok 4.7 cost | Claude Haiku 5.5 latency | Grok 4.7 latency |
|---|---|---|---|---|
| Vals Finance Index | $1.39 | $3.51 | 33m31s | 20m49s |
| Vals Index | $3.17 | $11.63 | 49m48s | 33m34s |
| Vals Legal Index | $2.35 | $5.82 | 58m22s | 24m33s |
| Vals Science Index | $1.86 | $7.12 | 39m58s | 40m34s |
| HLAB-Vals-1.1 | $4.26 | $10.75 | 43m41s | 35m48s |
| Legal Research Bench | $1.63 | $4.92 | 1h22m | 21m14s |
| EMB v1.1 | $1.36 | $6.48 | 19m56s | 34m57s |
| Finance Agent v2.1 | $1.64 | $2.24 | 31m05s | 10m52s |
| Tax Agent Bench | $1.16 | $1.80 | 49m34s | 16m36s |
| MedCode | N/A | N/A | 3m46s | 3m33s |
| MedScribe | N/A | N/A | 5m23s | 2m33s |
| ProofBench v1.1 | $0.17 | $0.79 | 6m58s | 12m47s |
| BioMysteryBench | $0.51 | $2.49 | 12m32s | 11m37s |
| MysteryMechanism | $0.54 | $1.84 | 13m25s | 17m33s |
| Terminal-Bench-Science | $4.52 | $17.04 | 1h34m | 1h33m |
| SAGE | N/A | N/A | 2m23s | 4m51s |
| Code Migration v1.1 | $6.20 | $38.05 | 1h26m | 1h05m |
| IOI | $0.83 | $12.71 | 12m34s | 46m48s |
| ProgramBench | $6.83 | $46.49 | 1h13m | 3h38m |
| SWE-bench Multimodal | $0.70 | $4.66 | 16m16s | 21m01s |
| Terminal-Bench 4.0 | $3.12 | $18.01 | 51m56s | 50m08s |
| Vibe Code Bench v1.1 | $6.07 | $15.83 | 34m36s | 35m47s |
| CyberBench v1.1 | $1.25 | $8.54 | 32m39s | 15m30s |
| SRE Bench | $3.14 | $64.77 | 57m56s | 1h41m |
| Public Benefits Bench v1.1 | $0.53 | $2.25 | 36m32s | 16m57s |
Results available only for Claude Haiku 5.5
- ClaimBench
- Vibe Code Bench 1-100
Results available only for Grok 4.7
- Vals RSI Index
- LegalBench
- Robo-bench