Model comparison
Claude Opus 4.7 vs Grok 4.7
Claude Opus 4.7 is the stronger model overall, scoring 58.3 to 53.1 on the Noometry Index. Grok 4.7 costs 3.3× less per token, which makes it the better buy when Claude Opus 4.7's lead doesn't matter for your workload.
Last verified . 37 shared benchmarks.
Summary
- They share 37 benchmarks with published results for both. Claude Opus 4.7 scores higher in 9 categories and Grok 4.7 in 1 category; 9 gaps are clear of the uncertainty.
- The widest gap is in agentic & tool use, where Claude Opus 4.7 leads 47.9 to 36.7.
- The biggest single-benchmark swing is NYT Connections (extended): 39% for Claude Opus 4.7 and 76.8% for Grok 4.7.
- Grok 4.7 is cheaper at $2 / $6 per million input/output tokens, against $5 / $25 for Claude Opus 4.7.
- Claude Opus 4.7 accepts more context: 1M tokens versus 500K.
Side by side
| Claude Opus 4.7 | Grok 4.7 | |
|---|---|---|
| Provider | Anthropic | xAI |
| Noometry Index | 58.3 | 53.1 |
| Released | 2026-04-14 | 2026-09-21 |
| Weights | Proprietary | Proprietary |
| Context window | 1M | 500K |
| Max output | 128K | 500K |
| Input $ / M tokens | $5 | $2 |
| Output $ / M tokens | $25 | $6 |
| Results tracked | 66 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Opus 4.7 leads
Claude Opus 4.7: 59.6 (#13), Grok 4.7: 58.0 (#18)
| Benchmark | Claude Opus 4.7 | Grok 4.7 |
|---|---|---|
| FrontierCode | 38.5% | 47.6% |
| LMArena WebDev | 1558 | 1639 |
| SciCode | 54.5% | 57.8% |
| LMArena Coding | 1518 | 1427 |
| SWE-bench Verified | 83.5% | — |
| CursorBench | — | 46.3% |
| FrontierSWE | — | 29.5% |
| GSO | 44.1% | — |
| WeirdML | 76.4% | — |
| MirrorCode | 31.1% | — |
| ALE-Bench | 1,323 | — |
Agentic & Tool Use Claude Opus 4.7 leads
Claude Opus 4.7: 47.9 (#10), Grok 4.7: 36.7 (#37)
| Benchmark | Claude Opus 4.7 | Grok 4.7 |
|---|---|---|
| APEX-Agents | 49.2% | 54.6% |
| GDP.pdf | 21% | 22.8% |
| Vending-Bench 2 | 10,937 | 10,537 |
| Terminal-Bench | 80.2% | — |
| OSWorld 2.0 | 18.2% | — |
| τ²-bench Banking | 40.2% | — |
| PostTrainBench | 28.6% | — |
| ExploitBench | 26.5% | — |
| GBAEval | 43.8% | — |
| LMArena Search | 1233 | — |
Reasoning Claude Opus 4.7 leads
Claude Opus 4.7: 53.8 (#29), Grok 4.7: 49.1 (#40)
| Benchmark | Claude Opus 4.7 | Grok 4.7 |
|---|---|---|
| NYT Connections (extended) | 39% | 76.8% |
| CritPt | 12% | 18% |
| Chess Puzzles | 30% | 38% |
| LMArena Hard Prompts | 1506 | 1413 |
| Mystery Game Puzzles | 28% | 29% |
| DTBench | 94.7% | 96% |
| LMCA | 52.2% | 49.4% |
| Epoch Capabilities Index | 156.25 | 153.53 |
| ARC-AGI-2 | 75.8% | — |
| SimpleBench | 61.7% | — |
| Kagi LLM Benchmark | 80.7% | — |
| ARC-AGI-1 | 93.5% | — |
| Thematic Generalization | 72.8% | — |
| EBR-Bench | 19% | — |
| ForecastBench | 60.3 | — |
Math Claude Opus 4.7 leads
Claude Opus 4.7: 66.7 (#26), Grok 4.7: 57.8 (#39)
| Benchmark | Claude Opus 4.7 | Grok 4.7 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 70.2% | 53% |
| FrontierMath Tier 4 | 31.7% | 17.1% |
| OTIS Mock AIME 2024-2025 | 97.8% | 98.1% |
| ProofBench | 54% | 34% |
| LMArena Math | 1499 | 1407 |
| MathArena Final-Answer Competitions | 73.6% | — |
| FrontierMath (Feb 2025 set) | 43.8% | — |
| FrontierMath Tier 4 (v1) | 22.9% | — |
Knowledge Too close to call
Claude Opus 4.7: 62.6 (#23), Grok 4.7: 62.8 (#22)
| Benchmark | Claude Opus 4.7 | Grok 4.7 |
|---|---|---|
| GPQA Diamond | 90.2% | 92.7% |
| SimpleQA Verified | 51.7% | 56% |
| LMArena Expert | 1521 | 1422 |
| Humanity's Last Exam | 36.2% | — |
| Vectara Hallucination Rate | 12% | — |
Multimodal Claude Opus 4.7 leads
Claude Opus 4.7: 41.2 (#38), Grok 4.7: 35.5 (#87)
| Benchmark | Claude Opus 4.7 | Grok 4.7 |
|---|---|---|
| LMArena Vision | 1316 | 1228 |
| Blueprint-Bench 2 | 24.5% | 32.5% |
| Furniture Assembly | 33.3% | 20.8% |
| LMArena Document | 1495 | — |
Multilingual Claude Opus 4.7 leads
Claude Opus 4.7: 57.3 (#10), Grok 4.7: 50.8 (#116)
| Benchmark | Claude Opus 4.7 | Grok 4.7 |
|---|---|---|
| LMArena Non-English | 1480 | 1389 |
| LMArena Chinese | 1531 | 1455 |
| LMArena French | 1503 | 1455 |
| LMArena Russian | 1494 | 1397 |
| LMArena Spanish | 1495 | 1400 |
| LMArena German | 1495 | — |
| LMArena Japanese | 1472 | — |
| LMArena Korean | 1464 | — |
Instruction Following Claude Opus 4.7 leads
Claude Opus 4.7: 78.4 (#10), Grok 4.7: 74.1 (#105)
| Benchmark | Claude Opus 4.7 | Grok 4.7 |
|---|---|---|
| LMArena Instruction Following | 1498 | 1404 |
Long Context Claude Opus 4.7 leads
Claude Opus 4.7: 46.2 (#25), Grok 4.7: 43.1 (#104)
| Benchmark | Claude Opus 4.7 | Grok 4.7 |
|---|---|---|
| LMArena Longer Query | 1505 | 1413 |
Writing & Preference Claude Opus 4.7 leads
Claude Opus 4.7: 75.1 (#8), Grok 4.7: 70.0 (#24)
| Benchmark | Claude Opus 4.7 | Grok 4.7 |
|---|---|---|
| LMArena Text | 1490 | 1399 |
| LMArena Creative Writing | 1486 | 1391 |
| EQ-Bench Creative Writing | 1914 | 2007 |
| LMArena Multi-Turn | 1505 | 1393 |
| EQ-Bench 4 | 1311 | — |
Frequently asked questions
Is Claude Opus 4.7 better than Grok 4.7?
Claude Opus 4.7 is the stronger model overall, scoring 58.3 to 53.1 on the Noometry Index. Grok 4.7 costs 3.3× less per token, which makes it the better buy when Claude Opus 4.7's lead doesn't matter for your workload.
Which is cheaper, Claude Opus 4.7 or Grok 4.7?
Grok 4.7 is cheaper. It lists at $2 per million input tokens and $6 per million output tokens; Claude Opus 4.7 lists at $5 and $25.
Is Claude Opus 4.7 or Grok 4.7 better for coding?
Claude Opus 4.7 scores higher on coding benchmarks: 59.6 versus 58.0 in the Noometry coding category.
Which has the bigger context window?
Claude Opus 4.7 does, with 1M tokens against 500K.
How many benchmarks do Claude Opus 4.7 and Grok 4.7 share?
37 benchmarks have published results for both models. Claude Opus 4.7 has 66 scored results on Noometry and Grok 4.7 has 39.