Grok 4.7 vs GPT-6 Sol: Benchmark Comparison
Grok 4.7 has the higher score on 6 of 18 shared benchmarks; GPT-6 Sol leads on 12.
The largest observed score gap is 57.00 pts on ProofBench v1.1 , where GPT-6 Sol leads.
Reported ±1 standard-error ranges overlap on 4 of 17 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 | Grok 4.7 | GPT-6 Sol | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals RSI Index | 20.20% | 28.11% | 7.91 pts | Uncertainty comparison unavailable |
| Vals Science Index | 34.83% ±1.61 | 45.00% ±2.26 | 10.17 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 47.12% ±3.47 | 28.85% ±3.15 | 18.27 pts | Reported ±1 SE ranges do not overlap |
| Tax Agent Bench | 35.75% ±3.46 | 15.00% ±2.58 | 20.75 pts | Reported ±1 SE ranges do not overlap |
| MedCode | 49.55% ±2.17 | 47.07% ±2.12 | 2.48 pts | Reported ±1 SE ranges overlap |
| MedScribe | 89.38% ±1.89 | 82.03% ±1.94 | 7.34 pts | Reported ±1 SE ranges do not overlap |
| ProofBench v1.1 | 26.00% ±4.41 | 83.00% ±3.77 | 57.00 pts | Reported ±1 SE ranges do not overlap |
| BioMysteryBench | 69.26% ±1.33 | 74.81% ±2.43 | 5.56 pts | Reported ±1 SE ranges do not overlap |
| MysteryMechanism | 25.23% ±2.92 | 30.18% ±3.09 | 4.95 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench-Science | 10.00% ±3.61 | 30.00% ±5.52 | 20.00 pts | Reported ±1 SE ranges do not overlap |
| IOI | 57.72% ±1.99 | 82.61% ±9.24 | 24.89 pts | Reported ±1 SE ranges do not overlap |
| ProgramBench | 0.50% ±0.50 | 2.00% ±0.99 | 1.50 pts | Reported ±1 SE ranges do not overlap |
| Terminal-Bench 4.0 | 28.79% ±2.31 | 44.44% ±3.54 | 15.66 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 86.17% ±2.18 | 87.82% ±2.53 | 1.65 pts | Reported ±1 SE ranges overlap |
| CyberBench v1.1 | 69.46% ±5.67 | 77.98% ±5.11 | 8.51 pts | Reported ±1 SE ranges overlap |
| SRE Bench | 3.82% ±1.19 | 30.53% ±2.85 | 26.72 pts | Reported ±1 SE ranges do not overlap |
| Financial Coaching Bench | 74.07% ±1.55 | 62.61% ±1.72 | 11.46 pts | Reported ±1 SE ranges do not overlap |
| Public Benefits Bench v1.1 | 65.63% ±1.24 | 56.63% ±1.29 | 9.00 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Grok 4.7 average | GPT-6 Sol average |
|---|---|---|
| Legal | 47.12% | 28.85% |
| Finance | 35.75% | 15.00% |
| Healthcare | 69.47% | 64.55% |
| Math | 26.00% | 83.00% |
| Science | 34.83% | 45.00% |
| Coding | 43.30% | 54.22% |
| Cyber | 36.64% | 54.25% |
| Social Mobility | 69.85% | 59.62% |
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 | Grok 4.7 cost | GPT-6 Sol cost | Grok 4.7 latency | GPT-6 Sol latency |
|---|---|---|---|---|
| Vals RSI Index | $237.12 | $665.11 | 90h00m | 90h00m |
| Vals Science Index | $7.12 | $2.31 | 40m34s | 31m16s |
| Legal Research Bench | $4.92 | $4.90 | 21m14s | 24m09s |
| Tax Agent Bench | $1.80 | $2.28 | 16m36s | 17m32s |
| MedCode | N/A | N/A | 3m33s | 62.71s |
| MedScribe | N/A | N/A | 2m33s | 81.14s |
| ProofBench v1.1 | $0.79 | $0.48 | 12m47s | 5m05s |
| BioMysteryBench | $2.49 | $0.66 | 11m37s | 5m24s |
| MysteryMechanism | $1.84 | $0.46 | 17m33s | 6m13s |
| Terminal-Bench-Science | $17.04 | $5.82 | 1h33m | 1h22m |
| IOI | $12.71 | $2.70 | 46m48s | 30m38s |
| ProgramBench | $46.49 | $11.67 | 3h38m | 46m11s |
| Terminal-Bench 4.0 | $18.01 | $5.79 | 50m08s | 34m56s |
| Vibe Code Bench v1.1 | $15.83 | $26.36 | 35m47s | 38m56s |
| CyberBench v1.1 | $8.54 | $1.34 | 15m30s | 12m14s |
| SRE Bench | $64.77 | $20.65 | 1h41m | 1h03m |
| Financial Coaching Bench | $0.24 | $0.23 | 5m12s | 6m51s |
| Public Benefits Bench v1.1 | $2.25 | $2.49 | 16m57s | 31m02s |
Results available only for Grok 4.7
- Vals Index
- Vals Finance Index
- Vals Legal Index
- HLAB-Vals-1.1
- Immigration
- LegalBench
- EMB v1.1
- Finance Agent v2.1
- Robo-bench
- SAGE
- Code Migration v1.1
- SWE-bench Multimodal
- Vibe Code Bench 1-100 v1.1
- CyberOffense
- Access to Justice Benchmark
Results available only for GPT-6 Sol
None.