Grok 4.7 vs Step 5 Preview
Grok 4.7 
Step 5 Preview Verdict
- Grok 4.7 (high) and Step 5 Preview are level on quality (65.8 vs 64.9).
- Grok 4.7 (high) is stronger in composite, knowledge.
- Step 5 Preview is stronger in long context, reasoning, multimodal.
- Step 5 Preview is 2.1× cheaper ($1.43 vs $3.00 per 1M blended).
- Step 5 Preview streams 1.7× faster (92 vs 55 tokens per second).
| Metric | Grok 4.7 (high) | Step 5 Preview |
|---|---|---|
| BenchLeader Index | 65.8 | 64.9 |
| Composite score | 87.5 | 84.2 |
| Knowledge score | 72.6 | 71.8 |
| Long context score | 64.8 | 70.7 |
| Reasoning score | 76.4 | 81.9 |
| Multimodal score | – | 60.8 |
| Blended price $/M | $3.00 | $1.43 |
| Output speed | 55 tok/s | 92 tok/s |
| Time to first answer | 1.4 s | 24.8 s |
| Context window | 500k | 1M |
| AA Intelligence Index | 46.3 | 43.7 |
| AA-LCR | 77.0% | 88.3% |
| MMMU-Pro | – | 76.4% |
| AA-Omniscience | 30.9 | 16.4 |
| Humanity's Last Exam (AA) | 42.3% | 46.5% |
| SciCode (AA) | 57.8% | 58.9% |
| LegalBench | 85.1% | – |
| CritPt | 18.0% | 20.9% |
| GDPval (AA) | 59.7% | 53.3% |
| Code Migration | 34.7% | – |
| Excel Modeling Benchmark | 54.9% | – |
| Harvey's Legal Agent Benchmark | 19.6% | – |
| Legal Research Bench | 39.9% | – |
| MedCode | 48.7% | – |
| MedScribe | 87.2% | – |
| ProgramBench | 0.0% | – |
| Public Benefits Bench | 68.5% | – |
| SAGE | 31.0% | – |
Data as of 2026-09-21. Best configuration of each model; every score links to its source on the model pages.
Grok 4.7 vs Step 5 Preview: questions
- Is Grok 4.7 better than Step 5 Preview?
- Grok 4.7 (high) and Step 5 Preview are level on quality (65.8 vs 64.9). The BenchLeader Index combines every independent quality benchmark; Grok 4.7 (high) is ahead overall as of 2026-09-21, but check the category scores for your use.
- Which is cheaper, Grok 4.7 or Step 5 Preview?
- Step 5 Preview is cheaper: $1.43 against $3.00 per million tokens, blended at three input tokens per output token.
- Which is faster, Grok 4.7 or Step 5 Preview?
- Step 5 Preview streams faster: 92 against 55 output tokens per second.
- Which has the larger context window?
- Step 5 Preview accepts more context: 1M against 500k tokens.