Claude Opus 5 vs Grok 4.20
Claude Opus 5 
Grok 4.20 Verdict
- Claude Opus 5 (high) leads on quality: 70.2 vs 59.9.
- Claude Opus 5 (high) is stronger in agents & tools, coding, composite, human preference, knowledge, long context, maths, multimodal, reasoning.
- Grok 4.20 (thinking) is stronger in instruction following.
- Grok 4.20 (thinking) is 6.4× cheaper ($1.56 vs $10.00 per 1M blended).
- Grok 4.20 (thinking) streams 1.9× faster (101 vs 54 tokens per second).
| Metric | Claude Opus 5 (high) | Grok 4.20 (thinking) |
|---|---|---|
| BenchLeader Index | 70.2 | 59.9 |
| Agents & tools score | 66.0 | 51.7 |
| Coding score | 71.9 | 54.7 |
| Composite score | 90.2 | 61.5 |
| Human preference score | 69.7 | 67.1 |
| Knowledge score | 80.3 | 56.8 |
| Long context score | 65.9 | 60.7 |
| Maths score | 72.3 | 55.6 |
| Multimodal score | 67.6 | 59.9 |
| Reasoning score | 75.6 | 60.7 |
| Instruction following score | – | 79.9 |
| Blended price $/M | $10.00 | $1.56 |
| Output speed | 54 tok/s | 101 tok/s |
| Time to first answer | 16.7 s | 21.4 s |
| Context window | 1M | 1M |
| GPQA Diamond | – | 89.3% |
| FrontierMath Tiers 1–3 | – | 44.9% |
| FrontierMath Tier 4 | – | 17.1% |
| OTIS Mock AIME | – | 92.2% |
| SimpleQA Verified | – | 30.2% |
| Terminal-Bench | 50.3% | – |
| OSWorld-Verified 2.0 | 29.0% | – |
| SciCode | 54.3% | – |
| WeirdML | 91.6% | – |
| ProofBench | – | 14.0% |
| LMArena Text | 1493 | 1472 |
| LMArena Hard Prompts | 1517 | 1488 |
| LMArena Coding | 1533 | 1510 |
| LMArena WebDev | 1660 | 1374 |
| LMArena Vision | 1321 | 1263 |
| LMArena Agent | 10.2 | – |
| AA Intelligence Index | 48.2 | 25.7 |
| IFBench | – | 82.9% |
| AA-LCR | 79.0% | 69.0% |
| MMMU-Pro | 82.4% | 74.6% |
| AA-Omniscience | 33.7 | 14.8 |
| Terminal-Bench Hard | – | 40.9% |
| GPQA Diamond (AA) | 93.7% | 91.1% |
| Humanity's Last Exam (AA) | 52.8% | 34.5% |
| SciCode (AA) | 55.4% | – |
| τ²-Bench Telecom (AA) | – | 96.5% |
| AIME (Vals) | – | 96.5% |
| LiveCodeBench | – | 84.3% |
| MMLU-Pro | – | 86.3% |
| LegalBench | – | 77.7% |
| CorpFin | – | 63.7% |
| TaxEval | – | 74.1% |
| MedQA | – | 94.5% |
| Terminal-Bench 2.1 (Vals) | – | 44.2% |
| SWE-bench (Vals) | – | 72.2% |
| GPQA Diamond (Vals) | – | 88.6% |
| Vals Index | – | 17.6 |
| Kagi LLM Benchmark | – | 75.0% |
| ARC-AGI-1 | 97.5% | 89.5% |
| ARC-AGI-2 | 88.3% | 65.1% |
| ARC-AGI-3 | 30.2% | 0.1% |
| CritPt | 28.3% | 6.6% |
| GDPval (AA) | 56.5% | – |
| τ²-Bench Banking (AA) | 44.7% | – |
| APEX-Agents (AA) | – | 14.2% |
| CaseLaw v2 | – | 54.5% |
| Code Migration | – | 0.3% |
| Excel Modeling Benchmark | – | 11.7% |
| Finance Agent v2 | – | 28.5% |
| Harvey's Legal Agent Benchmark | – | 0.0% |
| Legal Research Bench | – | 13.9% |
| MedCode | – | 32.2% |
| MedScribe | – | 63.4% |
| MMMU-Pro (Vals) | – | 83.5% |
| MortgageTax | – | 45.4% |
| SAGE | – | 38.2% |
| Terminal-Bench 2.0 (Vals) | – | 40.5% |
| Vals Multimodal Index | – | 39.1% |
| Vibe Code Bench v1.1 | – | 4.1% |
| LMArena Maths | 1525 | 1467 |
| LMArena Creative Writing | 1474 | 1446 |
| LMArena Instruction Following | 1499 | 1445 |
| LMArena Multi-turn | 1484 | 1480 |
| LMArena Longer Queries | 1508 | 1465 |
| LMArena Document | 1490 | 1439 |
| Chess Puzzles | – | 24.0% |
| DeepSWE | 72.8% | – |
| ForecastBench | – | 60.7% |
| CursorBench | 66.7% | – |
| ALE-Bench | 2164.6 | – |
Data as of 2026-09-19. Best configuration of each model; every score links to its source on the model pages.
Claude Opus 5 vs Grok 4.20: questions
- Is Claude Opus 5 better than Grok 4.20?
- Claude Opus 5 (high) leads on quality: 70.2 vs 59.9. The BenchLeader Index combines every independent quality benchmark; Claude Opus 5 (high) is ahead overall as of 2026-09-19, but check the category scores for your use.
- Is Claude Opus 5 better than Grok 4.20 for coding?
- Claude Opus 5 scores higher in coding (72 vs 55 on the category index, where 50 is average).
- Is Claude Opus 5 better than Grok 4.20 for agentic tasks?
- Claude Opus 5 scores higher in agentic tasks (66 vs 52 on the category index, where 50 is average).
- Which is cheaper, Claude Opus 5 or Grok 4.20?
- Grok 4.20 is cheaper: $1.56 against $10.00 per million tokens, blended at three input tokens per output token.
- Which is faster, Claude Opus 5 or Grok 4.20?
- Grok 4.20 streams faster: 101 against 54 output tokens per second.
- Which has the larger context window?
- Both accept 1M tokens of context.