Up to four models
Grok 4.7 vs Claude Sonnet 5
On the index (GPT-4 = 100), Grok 4.7 scores 164 and Claude Sonnet 5 164. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.
Grok 4.7
164149–180
WeightsClosed
ReleasedSep 2026
Price in / out$2.0 / $6.0
Context500K
Core tests taken6 of 9
Claude Sonnet 5
164148–180
WeightsClosed
ReleasedJun 2026
Price in / out$2.0 / $10
Context1M
Core tests taken5 of 9
Grok 4.7 and Claude Sonnet 5 are too close to call (0 points apart, within the uncertainty), ahead on 3 of 5 shared core tests, and costs 1.3× less.
Test by testthird-party results
Grok 4.7Claude Sonnet 5
Every result as a table
| Test | Grok 4.7 | Claude Sonnet 5 |
|---|---|---|
| Reasoning & knowledge | ||
| GPQA Diamond (Epoch AI) | 92.7% | 90.5% |
| GeoBench | 61.0% | 74.0% |
| SimpleQA Verified | 56.0% | 33.7% |
| Maths | ||
| FrontierMath (tiers 1-3) | 53.0% | 65.6% |
| AIME-style maths (OTIS mock) | 98.1% | 94.7% |
| ProofBench v1.1 | 26.0% | 77.0% |
| Chess puzzles | 41.0% | 45.0% |
| FrontierMath (tier 4) | 17.1% | 29.3% |
| Coding | ||
| Terminal-Bench 4.0 | 28.8% | 10.6% |
| Code Migration | 44.8% | 44.4% |
| CyberBench v1.1 | 69.5% | 70.3% |
| IOI | 57.7% | 45.0% |
| ProgramBench | 0.5% | 0.0% |
| Vibe Code Bench v1.1 | 86.2% | 81.3% |
| MirrorCode | 54.4% | 63.3% |
| Agents | ||
| APEX-Agents | 70.0% | 54.5% |
| SRE Bench | 3.8% | 0.4% |
| Terminal-Bench-Science | 10.0% | 5.7% |
| DTBench | 96.0% | 92.5% |
| LMCA | 49.4% | 50.0% |
| Mystery game puzzles | 29.0% | 35.0% |
| Professional work | ||
| Finance Agent (v2) | 52.3% | 53.9% |
| EMB | 67.0% | 66.3% |
| LegalBench | 84.4% | 83.9% |
| Legal Research Bench | 47.1% | 41.8% |
| MedCode | 49.5% | 47.5% |
| MedScribe | 89.4% | 76.0% |
| Public Benefits Bench v1.1 | 65.6% | 66.0% |
| Tax Agent Bench | 35.8% | 29.4% |
Tests at least two of these models have taken. Bold rows are the core tests.