Model comparison
Claude Opus 4.6 vs Grok 4.6
Claude Opus 4.6 is the stronger model overall, scoring 58.3 to 56.8 on the Noometry Index. Grok 4.6 costs 3.3× less per token, which makes it the better buy when Claude Opus 4.6's lead doesn't matter for your workload.
Last verified . 42 shared benchmarks.
Summary
- They share 42 benchmarks with published results for both. Claude Opus 4.6 scores higher in 5 categories and Grok 4.6 in 5 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in agentic & tool use, where Claude Opus 4.6 leads 51.0 to 39.4.
- The biggest single-benchmark swing is Chess Puzzles: 17% for Claude Opus 4.6 and 40% for Grok 4.6.
- Grok 4.6 is cheaper at $2 / $6 per million input/output tokens, against $5 / $25 for Claude Opus 4.6.
- Claude Opus 4.6 accepts more context: 1M tokens versus 500K.
Side by side
| Claude Opus 4.6 | Grok 4.6 | |
|---|---|---|
| Provider | Anthropic | xAI |
| Noometry Index | 58.3 | 56.8 |
| Released | 2026-02-04 | 2026-08-12 |
| Weights | Proprietary | Proprietary |
| Context window | 1M | 500K |
| Max output | 128K | 500K |
| Input $ / M tokens | $5 | $2 |
| Output $ / M tokens | $25 | $6 |
| Results tracked | 68 | 49 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Claude Opus 4.6: 57.7 (#15), Grok 4.6: 58.0 (#14)
| Benchmark | Claude Opus 4.6 | Grok 4.6 |
|---|---|---|
| FrontierCode | 26.6% | 48% |
| LMArena WebDev | 1546 | 1620 |
| WeirdML | 78% | 67.3% |
| LMArena Coding | 1535 | 1464 |
| ALE-Bench | 996.5 | 1,508 |
| SWE-bench Verified | 78.7% | — |
| DeepSWE | — | 67.5% |
| SWE-bench Verified (bash only) | 75.6% | — |
| CursorBench | — | 41.4% |
| SWE-bench Multilingual | 72% | — |
| FrontierSWE | — | 25.3% |
| SciCode | — | 56.5% |
| GSO | 41.2% | — |
| AlgoTune | 1.47 | — |
Agentic & Tool Use Claude Opus 4.6 leads
Claude Opus 4.6: 51.0 (#4), Grok 4.6: 39.4 (#25)
| Benchmark | Claude Opus 4.6 | Grok 4.6 |
|---|---|---|
| APEX-Agents | 46.3% | 65.3% |
| Vending-Bench 2 | 8,018 | 9,047 |
| Terminal-Bench | 79.8% | — |
| Remote Labor Index | 4.2% | — |
| τ²-bench Banking | 27.3% | — |
| Cybench | 93% | — |
| DeepResearch Bench | 55.3% | — |
| GBAEval | 44.1% | — |
| GDP.pdf | — | 17.2% |
| LMArena Search | 1253 | — |
| METR Time Horizons | 78.9% | — |
Reasoning Grok 4.6 leads
Claude Opus 4.6: 57.7 (#21), Grok 4.6: 61.3 (#18)
| Benchmark | Claude Opus 4.6 | Grok 4.6 |
|---|---|---|
| ARC-AGI-2 | 69.2% | 67.1% |
| SimpleBench | 67.6% | 75.9% |
| NYT Connections (extended) | 92.1% | 80% |
| ARC-AGI-1 | 94% | 87.5% |
| Chess Puzzles | 17% | 40% |
| EBR-Bench | 12.7% | 30.5% |
| LMArena Hard Prompts | 1527 | 1445 |
| Mystery Game Puzzles | 25% | 34% |
| DTBench | 91.2% | 97.3% |
| LMCA | 55.8% | 48.5% |
| Epoch Capabilities Index | 155.36 | 156.48 |
| Kagi LLM Benchmark | 83.6% | — |
| CritPt | — | 19.7% |
| EnigmaEval | 7.6% | — |
| Thematic Generalization | 80.6% | — |
| ForecastBench | 60 | — |
Math Grok 4.6 leads
Claude Opus 4.6: 62.7 (#25), Grok 4.6: 66.8 (#19)
| Benchmark | Claude Opus 4.6 | Grok 4.6 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 66% | 66% |
| FrontierMath Tier 4 | 26.8% | 31.7% |
| OTIS Mock AIME 2024-2025 | 94.4% | 99.2% |
| ProofBench | 50% | 51% |
| LMArena Math | 1517 | 1416 |
| MathArena Final-Answer Competitions | 78.5% | — |
| FrontierMath (Feb 2025 set) | 40.7% | — |
| FrontierMath Tier 4 (v1) | 22.9% | — |
Knowledge Grok 4.6 leads
Claude Opus 4.6: 61.8 (#23), Grok 4.6: 63.3 (#18)
| Benchmark | Claude Opus 4.6 | Grok 4.6 |
|---|---|---|
| GPQA Diamond | 90.5% | 94% |
| SimpleQA Verified | 47% | 49.3% |
| LMArena Expert | 1546 | 1463 |
| Humanity's Last Exam | 34.4% | — |
| Vectara Hallucination Rate | 12.2% | — |
Multimodal Grok 4.6 leads
Claude Opus 4.6: 37.0 (#67), Grok 4.6: 43.9 (#18)
| Benchmark | Claude Opus 4.6 | Grok 4.6 |
|---|---|---|
| LMArena Vision | 1316 | 1265 |
| Furniture Assembly | 28.3% | 40% |
| LMArena Document | 1507 | 1452 |
| Blueprint-Bench 2 | — | 33.2% |
Multilingual Claude Opus 4.6 leads
Claude Opus 4.6: 58.2 (#4), Grok 4.6: 53.3 (#67)
| Benchmark | Claude Opus 4.6 | Grok 4.6 |
|---|---|---|
| LMArena Non-English | 1490 | 1422 |
| LMArena Chinese | 1551 | 1479 |
| LMArena French | 1513 | 1458 |
| LMArena German | 1504 | 1425 |
| LMArena Japanese | 1484 | 1383 |
| LMArena Korean | 1465 | 1395 |
| LMArena Russian | 1497 | 1427 |
| LMArena Spanish | 1512 | 1408 |
Instruction Following Claude Opus 4.6 leads
Claude Opus 4.6: 79.7 (#3), Grok 4.6: 75.6 (#59)
| Benchmark | Claude Opus 4.6 | Grok 4.6 |
|---|---|---|
| LMArena Instruction Following | 1523 | 1431 |
Long Context Claude Opus 4.6 leads
Claude Opus 4.6: 48.1 (#13), Grok 4.6: 44.4 (#62)
| Benchmark | Claude Opus 4.6 | Grok 4.6 |
|---|---|---|
| LMArena Longer Query | 1520 | 1453 |
| CL-bench | 20.7% | — |
| CL-bench Life | 17% | — |
Writing & Preference Claude Opus 4.6 leads
Claude Opus 4.6: 73.7 (#9), Grok 4.6: 62.5 (#77)
| Benchmark | Claude Opus 4.6 | Grok 4.6 |
|---|---|---|
| LMArena Text | 1504 | 1427 |
| LMArena Creative Writing | 1506 | 1430 |
| LMArena Multi-Turn | 1514 | 1422 |
| EQ-Bench Creative Writing | 1809 | — |
| EQ-Bench 4 | 1223 | — |
Frequently asked questions
Is Claude Opus 4.6 better than Grok 4.6?
Claude Opus 4.6 is the stronger model overall, scoring 58.3 to 56.8 on the Noometry Index. Grok 4.6 costs 3.3× less per token, which makes it the better buy when Claude Opus 4.6's lead doesn't matter for your workload.
Which is cheaper, Claude Opus 4.6 or Grok 4.6?
Grok 4.6 is cheaper. It lists at $2 per million input tokens and $6 per million output tokens; Claude Opus 4.6 lists at $5 and $25.
Is Claude Opus 4.6 or Grok 4.6 better for coding?
They score almost the same on coding (57.7 vs 58.0); test both on your own repository before choosing.
Which has the bigger context window?
Claude Opus 4.6 does, with 1M tokens against 500K.
How many benchmarks do Claude Opus 4.6 and Grok 4.6 share?
42 benchmarks have published results for both models. Claude Opus 4.6 has 68 scored results on Noometry and Grok 4.6 has 49.