Claude Opus 4.8 vs Grok 4.6: Benchmark Comparison
Claude Opus 4.8 has the higher score on 14 of 25 shared benchmarks; Grok 4.6 leads on 10.
The largest observed score gap is 25.89 pts on SAGE , where Claude Opus 4.8 leads.
Reported ±1 standard-error ranges overlap on 12 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 | Claude Opus 4.8 | Grok 4.6 | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Index | 55.10% ±1.01 | 52.09% ±1.15 | 3.01 pts | Reported ±1 SE ranges do not overlap |
| Vals RSI Index | 23.98% | 25.46% | 1.48 pts | Uncertainty comparison unavailable |
| Harvey's Legal Agent Benchmark | 9.58% ±2.15 | 15.83% ±2.65 | 6.25 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 43.75% ±3.45 | 48.08% ±3.47 | 4.33 pts | Reported ±1 SE ranges overlap |
| LegalBench | 83.57% ±0.47 | 86.31% ±0.42 | 2.74 pts | Reported ±1 SE ranges do not overlap |
| EMB | 69.37% ±2.61 | 62.73% ±3.08 | 6.65 pts | Reported ±1 SE ranges do not overlap |
| Finance Agent (v2) | 53.92% ±0.16 | 53.68% ±0.67 | 0.24 pts | Reported ±1 SE ranges overlap |
| MortgageTax | 69.91% ±0.89 | 64.19% ±0.95 | 5.72 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 64.91% ±3.17 | 70.79% ±3.00 | 5.88 pts | Reported ±1 SE ranges overlap |
| TaxEval v2 | 75.63% ±0.84 | 71.10% ±0.90 | 4.54 pts | Reported ±1 SE ranges do not overlap |
| MedCode | 53.22% ±2.17 | 44.71% ±2.26 | 8.50 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 85.75% ±1.93 | 86.53% ±1.96 | 0.78 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench Science | 4.29% ±2.44 | 11.43% ±3.83 | 7.14 pts | Reported ±1 SE ranges do not overlap |
| GPQA Diamond | 92.42% ±1.98 | 94.70% ±1.13 | 2.27 pts | Reported ±1 SE ranges overlap |
| MMLU Pro | 89.58% ±0.30 | 89.40% ±0.30 | 0.18 pts | Reported ±1 SE ranges overlap |
| SAGE | 54.79% ±3.35 | 28.90% ±3.08 | 25.89 pts | Reported ±1 SE ranges do not overlap |
| Code Migration | 47.25% ±4.18 | 44.57% ±4.48 | 2.68 pts | Reported ±1 SE ranges overlap |
| LiveCodeBench | 87.82% ±0.95 | 88.22% ±0.94 | 0.41 pts | Reported ±1 SE ranges overlap |
| ProgramBench | 1.00% ±0.70 | 1.00% ±0.70 | 0.00 pts | Reported ±1 SE ranges overlap |
| SkillsBench | 59.23% ±4.50 | 55.77% ±4.84 | 3.46 pts | Reported ±1 SE ranges overlap |
| SWE-bench | 88.60% ±1.42 | 95.60% ±0.92 | 7.00 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench 4.0 | 23.23% ±1.34 | 17.17% ±1.34 | 6.06 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 82.72% ±3.08 | 76.24% ±3.82 | 6.49 pts | Reported ±1 SE ranges overlap |
| Public Benefits Bench v1.1 | 68.13% ±1.21 | 66.85% ±1.23 | 1.29 pts | Reported ±1 SE ranges overlap |
| Time Horizon Index: KSP | 11.83% ±0.00 | 5.83% ±0.00 | 6.00 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Claude Opus 4.8 average | Grok 4.6 average |
|---|---|---|
| Legal | 45.63% | 50.07% |
| Finance | 66.75% | 64.50% |
| Healthcare | 69.49% | 65.62% |
| Science | 4.29% | 11.43% |
| Academic | 91.00% | 92.05% |
| Education | 54.79% | 28.90% |
| Coding | 55.69% | 54.08% |
| Social Mobility | 68.13% | 66.85% |
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 4.8 cost | Grok 4.6 cost | Claude Opus 4.8 latency | Grok 4.6 latency |
|---|---|---|---|---|
| Vals Index | $13.14 | $4.49 | 39m48s | 37m01s |
| Vals RSI Index | $465.25 | $88.12 | 90h00m | 90h00m |
| Harvey's Legal Agent Benchmark | $10.22 | $4.01 | 24m08s | 45m05s |
| Legal Research Bench | $2.82 | $1.53 | 11m54s | 25m04s |
| LegalBench | N/A | N/A | 3.11s | 27.42s |
| EMB | $12.06 | $3.06 | 43m52s | 52m59s |
| Finance Agent (v2) | $4.22 | $1.66 | 9m07s | 16m41s |
| MortgageTax | N/A | N/A | 20.81s | 8.55s |
| Tax Agent Bench | $1.69 | $0.98 | 7m59s | 10m58s |
| TaxEval v2 | N/A | N/A | 70.33s | 47.57s |
| MedCode | N/A | N/A | 105.92s | 2m07s |
| MedScribe | N/A | N/A | 82.36s | 70.94s |
| Terminal-Bench Science | $23.14 | $4.82 | 1h55m | 1h18m |
| GPQA Diamond | N/A | N/A | 2m14s | 3m29s |
| MMLU Pro | N/A | N/A | 23.44s | 54.98s |
| SAGE | N/A | N/A | 3m06s | 3m22s |
| Code Migration | $30.51 | $14.58 | 1h15m | 1h30m |
| LiveCodeBench | N/A | N/A | 2m40s | 118.40s |
| ProgramBench | $31.27 | $16.76 | 1h32m | 4h27m |
| SkillsBench | $4.77 | $0.92 | 15m30s | 13m28s |
| SWE-bench | $1.92 | $0.78 | 9m27s | 10m04s |
| Terminal-Bench 4.0 | $17.14 | $5.07 | 1h11m | 29m27s |
| Vibe Code Bench v1.1 | $26.88 | $4.88 | 1h16m | 25m30s |
| Public Benefits Bench v1.1 | $1.89 | $0.92 | 12m48s | 25m53s |
| Time Horizon Index: KSP | $1499.32 | $225.12 | N/A | N/A |
Results available only for Claude Opus 4.8
- MMMU Pro
Results available only for Grok 4.6
- ProofBench v1.1
- BioMysteryBench
- MysteryMechanism
- IOI
- Vibe Code Bench 1-100
- CyberBench v1.1