Grok 4.7 vs GPT-5.6 Terra: Benchmark Comparison
Grok 4.7 has the higher score on 10 of 19 shared benchmarks; GPT-5.6 Terra leads on 7.
The largest observed score gap is 48.00 pts on ProofBench v1.1 , where GPT-5.6 Terra leads.
Reported ±1 standard-error ranges overlap on 11 of 19 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-5.6 Terra | Gap | Reported uncertainty |
|---|---|---|---|---|
| Vals Index | 54.95% ±1.07 | 53.09% ±1.29 | 1.86 pts | Reported ±1 SE ranges overlap |
| Harvey's Legal Agent Benchmark | 12.50% ±2.66 | 0.83% ±0.83 | 11.67 pts | Reported ±1 SE ranges do not overlap |
| Legal Research Bench | 47.12% ±3.47 | 41.35% ±3.42 | 5.77 pts | Reported ±1 SE ranges overlap |
| LegalBench | 84.39% ±0.46 | 85.11% ±0.45 | 0.72 pts | Reported ±1 SE ranges overlap |
| EMB | 66.99% ±3.04 | 66.20% ±3.11 | 0.79 pts | Reported ±1 SE ranges overlap |
| Finance Agent (v2) | 52.25% ±0.35 | 54.44% ±2.07 | 2.19 pts | Reported ±1 SE ranges overlap |
| Tax Agent Bench | 65.60% ±3.24 | 65.20% ±3.19 | 0.40 pts | Reported ±1 SE ranges overlap |
| MedCode | 49.55% ±2.17 | 43.41% ±2.17 | 6.14 pts | Reported ±1 SE ranges do not overlap |
| MedScribe | 89.38% ±1.89 | 82.87% ±1.95 | 6.51 pts | Reported ±1 SE ranges do not overlap |
| ProofBench v1.1 | 26.00% ±4.41 | 74.00% ±4.41 | 48.00 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 | 40.79% ±3.35 | 47.00% ±3.40 | 6.21 pts | Reported ±1 SE ranges overlap |
| Code Migration | 44.82% ±4.21 | 47.80% ±4.28 | 2.99 pts | Reported ±1 SE ranges overlap |
| IOI | 57.72% ±1.99 | 87.61% ±6.53 | 29.89 pts | Reported ±1 SE ranges do not overlap |
| ProgramBench | 0.50% ±0.50 | 0.50% ±0.50 | 0.00 pts | Reported ±1 SE ranges overlap |
| Terminal-Bench 4.0 | 28.79% ±2.31 | 22.73% ±2.31 | 6.06 pts | Reported ±1 SE ranges do not overlap |
| Vibe Code Bench v1.1 | 86.17% ±2.18 | 74.59% ±4.20 | 11.58 pts | Reported ±1 SE ranges do not overlap |
| CyberBench v1.1 | 69.46% ±5.67 | 72.08% ±5.41 | 2.62 pts | Reported ±1 SE ranges overlap |
| Public Benefits Bench v1.1 | 65.63% ±1.24 | 62.38% ±1.26 | 3.25 pts | Reported ±1 SE ranges do not overlap |
Performance by category
| Category | Grok 4.7 average | GPT-5.6 Terra average |
|---|---|---|
| Legal | 48.00% | 42.43% |
| Finance | 61.61% | 61.95% |
| Healthcare | 69.47% | 63.14% |
| Math | 26.00% | 74.00% |
| Science | 10.00% | 10.00% |
| Education | 40.79% | 47.00% |
| Coding | 43.60% | 46.65% |
| Cyber | 69.46% | 72.08% |
| Social Mobility | 65.63% | 62.38% |
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-5.6 Terra cost | Grok 4.7 latency | GPT-5.6 Terra latency |
|---|---|---|---|---|
| Vals Index | $12.12 | $5.78 | 35m31s | 35m39s |
| Harvey's Legal Agent Benchmark | $11.13 | $3.20 | 42m05s | 17m05s |
| Legal Research Bench | $4.92 | $7.91 | 21m14s | 1h10m |
| LegalBench | N/A | N/A | 26.02s | 2.78s |
| EMB | $6.48 | $2.19 | 34m57s | 13m33s |
| Finance Agent (v2) | $2.72 | $3.63 | 15m33s | 24m10s |
| Tax Agent Bench | $1.80 | $4.83 | 16m36s | 1h01m |
| MedCode | N/A | N/A | 3m33s | 18.41s |
| MedScribe | N/A | N/A | 2m33s | 35.53s |
| ProofBench v1.1 | $0.79 | $1.27 | 12m47s | 12m57s |
| Terminal-Bench Science | $14.45 | $5.19 | 1h29m | 2h04m |
| SAGE | N/A | N/A | 4m51s | 40.06s |
| Code Migration | $36.55 | $8.13 | 1h09m | 48m34s |
| IOI | $12.71 | $8.66 | 46m48s | 1h23m |
| ProgramBench | $46.49 | $5.38 | 3h38m | 33m36s |
| Terminal-Bench 4.0 | $18.09 | $5.60 | 50m21s | 31m29s |
| Vibe Code Bench v1.1 | $15.83 | $7.89 | 35m47s | 19m12s |
| CyberBench v1.1 | $8.54 | $3.31 | 15m30s | 17m04s |
| Public Benefits Bench v1.1 | $2.25 | $1.20 | 16m57s | 19m31s |
Results available only for Grok 4.7
- Vals RSI Index
- BioMysteryBench
- MysteryMechanism
Results available only for GPT-5.6 Terra
- MortgageTax
- TaxEval v2
- GPQA Diamond
- MMLU Pro
- MMMU Pro
- LiveCodeBench
- SkillsBench
- SWE-bench
- Vibe Code Bench 1-100