Claude Opus 5.5 vs Grok 4.7: Benchmark Comparison
Claude Opus 5.5 has the higher score on 19 of 21 shared benchmarks; Grok 4.7 leads on 2.
The largest observed score gap is 74.00 pts on ProofBench v1.1 , where Claude Opus 5.5 leads.
Reported ±1 standard-error ranges overlap on 5 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 | Claude Opus 5.5 | Grok 4.7 | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Index | 66.97% ±0.89 | 54.95% ±1.07 | 12.02 pts | Reported ±1 SE ranges do not overlap |
| Vals RSI Index | 37.31% | 20.20% | 17.11 pts | Uncertainty comparison unavailable |
| Harvey's Legal Agent Benchmark | 3.75% ±1.47 | 12.50% ±2.66 | 8.75 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 50.48% ±3.48 | 47.12% ±3.47 | 3.37 pts | Reported ±1 SE ranges overlap |
| EMB | 75.94% ±2.38 | 66.99% ±3.04 | 8.95 pts | Reported ±1 SE ranges do not overlap |
| Finance Agent (v2) | 58.59% ±0.17 | 52.25% ±0.35 | 6.34 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 70.50% ±3.15 | 65.60% ±3.24 | 4.90 pts | Reported ±1 SE ranges overlap |
| MedCode | 49.80% ±2.27 | 49.55% ±2.17 | 0.24 pts | Reported ±1 SE ranges overlap |
| MedScribe | 91.43% ±1.93 | 89.38% ±1.89 | 2.05 pts | Reported ±1 SE ranges overlap |
| ProofBench v1.1 | 100.00% ±0.00 | 26.00% ±4.41 | 74.00 pts | Reported ±1 SE ranges do not overlap |
| BioMysteryBench | 79.26% ±0.98 | 69.26% ±1.33 | 10.00 pts | Reported ±1 SE ranges do not overlap |
| MysteryMechanism | 49.55% ±3.36 | 25.23% ±2.92 | 24.32 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench Science | 47.14% ±6.01 | 10.00% ±3.61 | 37.14 pts | Reported ±1 SE ranges do not overlap |
| SAGE | 45.83% ±3.36 | 40.79% ±3.35 | 5.04 pts | Reported ±1 SE ranges overlap |
| Code Migration | 66.65% ±4.33 | 44.82% ±4.21 | 21.83 pts | Reported ±1 SE ranges do not overlap |
| IOI | 95.06% ±4.94 | 57.72% ±1.99 | 37.33 pts | Reported ±1 SE ranges do not overlap |
| ProgramBench | 18.50% ±2.75 | 0.50% ±0.50 | 18.00 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench 4.0 | 65.15% ±0.00 | 28.79% ±2.31 | 36.36 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 90.29% ±1.53 | 86.17% ±2.18 | 4.12 pts | Reported ±1 SE ranges do not overlap |
| CyberBench v1.1 | 55.36% ±5.13 | 69.46% ±5.67 | 14.11 pts | Reported ±1 SE ranges do not overlap |
| Public Benefits Bench v1.1 | 70.64% ±1.19 | 65.63% ±1.24 | 5.01 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Claude Opus 5.5 average | Grok 4.7 average |
|---|---|---|
| Legal | 27.12% | 29.81% |
| Finance | 68.34% | 61.61% |
| Healthcare | 70.61% | 69.47% |
| Math | 100.00% | 26.00% |
| Science | 58.65% | 34.83% |
| Education | 45.83% | 40.79% |
| Coding | 67.13% | 43.60% |
| Cyber | 55.36% | 69.46% |
| Social Mobility | 70.64% | 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 Opus 5.5 cost | Grok 4.7 cost | Claude Opus 5.5 latency | Grok 4.7 latency |
|---|---|---|---|---|
| Vals Index | $32.14 | $12.12 | 1h19m | 35m31s |
| Vals RSI Index | $411.86 | $237.12 | 90h00m | 90h00m |
| Harvey's Legal Agent Benchmark | $21.38 | $11.13 | 49m00s | 42m05s |
| Legal Research Bench | $25.16 | $4.92 | 1h52m | 21m14s |
| EMB | $11.02 | $6.48 | 35m06s | 34m57s |
| Finance Agent (v2) | $9.22 | $2.72 | 33m08s | 15m33s |
| Tax Agent Bench | $15.09 | $1.80 | 1h07m | 16m36s |
| MedCode | N/A | N/A | 4m07s | 3m33s |
| MedScribe | N/A | N/A | 7m22s | 2m33s |
| ProofBench v1.1 | $0.96 | $0.79 | 5m51s | 12m47s |
| BioMysteryBench | $3.64 | $2.49 | 12m19s | 11m37s |
| MysteryMechanism | $4.62 | $1.84 | 19m38s | 17m33s |
| Terminal-Bench Science | $19.12 | $14.45 | 2h28m | 1h29m |
| SAGE | N/A | N/A | 114.20s | 4m51s |
| Code Migration | $112.97 | $36.55 | 3h19m | 1h09m |
| IOI | $5.26 | $12.71 | 17m15s | 46m48s |
| ProgramBench | $68.05 | $46.49 | 2h24m | 3h38m |
| Terminal-Bench 4.0 | $13.20 | $18.09 | 1h04m | 50m21s |
| Vibe Code Bench v1.1 | $57.92 | $15.83 | 1h36m | 35m47s |
| CyberBench v1.1 | $3.48 | $8.54 | 18m11s | 15m30s |
| Public Benefits Bench v1.1 | $5.95 | $2.25 | 1h20m | 16m57s |
Results available only for Claude Opus 5.5
- Vibe Code Bench 1-100
- SRE Bench
- CUA-bench
- Time Horizon Index: KSP
Results available only for Grok 4.7
- LegalBench