Model comparison
Claude Sonnet 5 vs Grok 4.6
Grok 4.6 is the stronger model overall, scoring 56.9 to 54.6 on the Noometry Index.
Last verified . 43 shared benchmarks.
Summary
- They share 43 benchmarks with published results for both. Claude Sonnet 5 scores higher in 5 categories and Grok 4.6 in 5 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Grok 4.6 leads 61.4 to 49.1.
- The biggest single-benchmark swing is ProofBench: 77% for Claude Sonnet 5 and 51% for Grok 4.6.
- Grok 4.6 is cheaper at $2 / $6 per million input/output tokens, against $2 / $10 for Claude Sonnet 5.
- Claude Sonnet 5 accepts more context: 1M tokens versus 500K.
Side by side
| Claude Sonnet 5 | Grok 4.6 | |
|---|---|---|
| Provider | Anthropic | xAI |
| Noometry Index | 54.6 | 56.9 |
| Released | 2026-06-29 | 2026-08-12 |
| Weights | Proprietary | Proprietary |
| Context window | 1M | 500K |
| Max output | 128K | 500K |
| Input $ / M tokens | $2 | $2 |
| Output $ / M tokens | $10 | $6 |
| Results tracked | 51 | 49 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4.6 leads
Claude Sonnet 5: 55.5 (#26), Grok 4.6: 58.5 (#16)
| Benchmark | Claude Sonnet 5 | Grok 4.6 |
|---|---|---|
| DeepSWE | 53.8% | 67.5% |
| FrontierCode | 42.7% | 48% |
| CursorBench | 34.1% | 41.4% |
| LMArena WebDev | 1541 | 1617 |
| SciCode | 54.3% | 56.5% |
| WeirdML | 68.8% | 67.3% |
| LMArena Coding | 1483 | 1465 |
| ALE-Bench | 1,463 | 1,508 |
| FrontierSWE | — | 25.3% |
| GSO | 37.3% | — |
Agentic & Tool Use Claude Sonnet 5 leads
Claude Sonnet 5: 42.8 (#18), Grok 4.6: 39.4 (#27)
| Benchmark | Claude Sonnet 5 | Grok 4.6 |
|---|---|---|
| APEX-Agents | 54.5% | 65.3% |
| Vending-Bench 2 | 6,378 | 9,047 |
| GBAEval | 65.3% | — |
| GDP.pdf | — | 17.2% |
| LMArena Search | 1194 | — |
Reasoning Grok 4.6 leads
Claude Sonnet 5: 49.1 (#39), Grok 4.6: 61.4 (#20)
| Benchmark | Claude Sonnet 5 | Grok 4.6 |
|---|---|---|
| SimpleBench | 60.6% | 75.9% |
| NYT Connections (extended) | 75.1% | 80% |
| CritPt | 16.9% | 19.7% |
| Chess Puzzles | 35% | 40% |
| LMArena Hard Prompts | 1461 | 1447 |
| Mystery Game Puzzles | 35% | 34% |
| DTBench | 92.5% | 97.3% |
| LMCA | 50% | 48.5% |
| Epoch Capabilities Index | 156.21 | 156.44 |
| ARC-AGI-2 | — | 67.1% |
| ARC-AGI-1 | — | 87.5% |
| EBR-Bench | — | 30.5% |
| Surface Evolver Bench | 60% | — |
| Bench to the Future 3 | 0.14 | — |
| ForecastBench | 61.1 | — |
Math Too close to call
Claude Sonnet 5: 66.2 (#27), Grok 4.6: 67.0 (#24)
| Benchmark | Claude Sonnet 5 | Grok 4.6 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 65.6% | 66% |
| FrontierMath Tier 4 | 29.3% | 31.7% |
| OTIS Mock AIME 2024-2025 | 94.7% | 99.2% |
| ProofBench | 77% | 51% |
| LMArena Math | 1467 | 1423 |
Knowledge Grok 4.6 leads
Claude Sonnet 5: 55.6 (#47), Grok 4.6: 63.3 (#20)
| Benchmark | Claude Sonnet 5 | Grok 4.6 |
|---|---|---|
| GPQA Diamond | 90.5% | 94% |
| SimpleQA Verified | 33.7% | 49.3% |
| LMArena Expert | 1490 | 1467 |
Multimodal Grok 4.6 leads
Claude Sonnet 5: 42.4 (#31), Grok 4.6: 43.6 (#23)
| Benchmark | Claude Sonnet 5 | Grok 4.6 |
|---|---|---|
| LMArena Vision | 1274 | 1263 |
| Blueprint-Bench 2 | 24.9% | 33.2% |
| LMArena Document | 1466 | 1452 |
| Furniture Assembly | — | 40% |
Multilingual Too close to call
Claude Sonnet 5: 53.8 (#55), Grok 4.6: 53.0 (#74)
| Benchmark | Claude Sonnet 5 | Grok 4.6 |
|---|---|---|
| LMArena Non-English | 1431 | 1420 |
| LMArena Chinese | 1477 | 1480 |
| LMArena French | 1460 | 1461 |
| LMArena German | 1440 | 1431 |
| LMArena Japanese | 1422 | 1376 |
| LMArena Korean | 1411 | 1397 |
| LMArena Russian | 1451 | 1422 |
| LMArena Spanish | 1437 | 1404 |
Instruction Following Too close to call
Claude Sonnet 5: 76.3 (#41), Grok 4.6: 75.4 (#63)
| Benchmark | Claude Sonnet 5 | Grok 4.6 |
|---|---|---|
| LMArena Instruction Following | 1452 | 1431 |
Long Context Too close to call
Claude Sonnet 5: 44.8 (#55), Grok 4.6: 44.5 (#66)
| Benchmark | Claude Sonnet 5 | Grok 4.6 |
|---|---|---|
| LMArena Longer Query | 1463 | 1454 |
Writing & Preference Claude Sonnet 5 leads
Claude Sonnet 5: 69.2 (#25), Grok 4.6: 62.3 (#80)
| Benchmark | Claude Sonnet 5 | Grok 4.6 |
|---|---|---|
| LMArena Text | 1442 | 1428 |
| LMArena Creative Writing | 1416 | 1428 |
| LMArena Multi-Turn | 1454 | 1425 |
| EQ-Bench Creative Writing | 1794 | — |
| EQ-Bench 4 | 1236 | — |
Frequently asked questions
Is Claude Sonnet 5 better than Grok 4.6?
Grok 4.6 is the stronger model overall, scoring 56.9 to 54.6 on the Noometry Index.
Which is cheaper, Claude Sonnet 5 or Grok 4.6?
Grok 4.6 is cheaper. It lists at $2 per million input tokens and $6 per million output tokens; Claude Sonnet 5 lists at $2 and $10.
Is Claude Sonnet 5 or Grok 4.6 better for coding?
Grok 4.6 scores higher on coding benchmarks: 58.5 versus 55.5 in the Noometry coding category.
Which has the bigger context window?
Claude Sonnet 5 does, with 1M tokens against 500K.
How many benchmarks do Claude Sonnet 5 and Grok 4.6 share?
43 benchmarks have published results for both models. Claude Sonnet 5 has 51 scored results on Noometry and Grok 4.6 has 49.