GPT-5.6 Sol vs Grok 4.6
GPT-5.6 Sol 
Grok 4.6 Verdict
- GPT-5.6 Sol (high) leads on quality: 68.0 vs 63.8.
- GPT-5.6 Sol (high) is stronger in agents & tools, coding, instruction following, long context, multimodal, reasoning.
- Grok 4.6 (medium) is stronger in composite, knowledge.
- Grok 4.6 (medium) is 2.7× cheaper ($3.00 vs $8.00 per 1M blended).
| Metric | GPT-5.6 Sol (high) | Grok 4.6 (medium) |
|---|---|---|
| BenchLeader Index | 68.0 | 63.8 |
| Agents & tools score | 87.6 | – |
| Coding score | 72.0 | 63.9 |
| Composite score | 82.5 | 83.2 |
| Instruction following score | 67.6 | – |
| Knowledge score | 73.7 | 77.3 |
| Long context score | 67.4 | 67.1 |
| Multimodal score | 66.5 | – |
| Reasoning score | 62.9 | 62.1 |
| Blended price $/M | $8.00 | $3.00 |
| Output speed | 59 tok/s | 55 tok/s |
| Time to first answer | 31.4 s | 36.7 s |
| Context window | 1M | 500k |
| SciCode | 56.9% | 54.6% |
| WeirdML | 88.8% | – |
| AA Intelligence Index | 42.5 | 43.0 |
| IFBench | 69.2% | – |
| AA-LCR | 81.7% | 81.0% |
| MMMU-Pro | 81.8% | – |
| AA-Omniscience | 20.4 | 28 |
| Terminal-Bench Hard | 62.1% | – |
| GPQA Diamond (AA) | 92.8% | 93.5% |
| Humanity's Last Exam (AA) | 46.0% | 42.1% |
| SciCode (AA) | 57.8% | 55.9% |
| τ²-Bench Telecom (AA) | 83.3% | – |
| EnigmaEval | 37.1% | – |
| ARC-AGI-1 | 97.0% | 87.5% |
| ARC-AGI-2 | 85.4% | 61.3% |
| ARC-AGI-3 | 2.2% | – |
Data as of 2026-09-09. Best configuration of each model; every score links to its source on the model pages.