Model comparison
Claude Sonnet 4.5 vs Grok 4.6
Grok 4.6 is the stronger model overall, scoring 56.9 to 44.1 on the Noometry Index.
Last verified . 40 shared benchmarks.
Summary
- They share 40 benchmarks with published results for both. Claude Sonnet 4.5 scores higher in 3 categories and Grok 4.6 in 7 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Grok 4.6 leads 67.0 to 32.3.
- The biggest single-benchmark swing is ARC-AGI-2: 13.6% for Claude Sonnet 4.5 and 67.1% for Grok 4.6.
- Grok 4.6 is cheaper at $2 / $6 per million input/output tokens, against $3 / $15 for Claude Sonnet 4.5.
- Grok 4.6 accepts more context: 500K tokens versus 200K.
Side by side
| Claude Sonnet 4.5 | Grok 4.6 | |
|---|---|---|
| Provider | Anthropic | xAI |
| Noometry Index | 44.1 | 56.9 |
| Released | 2025-09-29 | 2026-08-12 |
| Weights | Proprietary | Proprietary |
| Context window | 200K | 500K |
| Max output | 64K | 500K |
| Input $ / M tokens | $3 | $2 |
| Output $ / M tokens | $15 | $6 |
| Results tracked | 73 | 49 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4.6 leads
Claude Sonnet 4.5: 47.3 (#61), Grok 4.6: 58.5 (#16)
| Benchmark | Claude Sonnet 4.5 | Grok 4.6 |
|---|---|---|
| LMArena WebDev | 1393 | 1617 |
| SciCode | 44.7% | 56.5% |
| WeirdML | 47.7% | 67.3% |
| LMArena Coding | 1489 | 1465 |
| ALE-Bench | 796.15 | 1,508 |
| SWE-bench Verified | 71.3% | — |
| DeepSWE | — | 67.5% |
| FrontierCode | — | 48% |
| SWE-bench Verified (bash only) | 71.4% | — |
| CursorBench | — | 41.4% |
| SWE-bench Multilingual | 67% | — |
| FrontierSWE | — | 25.3% |
| GSO | 14.7% | — |
| AlgoTune | 1.52 | — |
Agentic & Tool Use Grok 4.6 leads
Claude Sonnet 4.5: 38.3 (#32), Grok 4.6: 39.4 (#27)
| Benchmark | Claude Sonnet 4.5 | Grok 4.6 |
|---|---|---|
| Vending-Bench 2 | 3,839 | 9,047 |
| Terminal-Bench | 46.5% | — |
| APEX-Agents | — | 65.3% |
| Berkeley Function Calling Leaderboard | 73.2% | — |
| GDPval | 42.5% | — |
| Remote Labor Index | 2.1% | — |
| τ²-bench Airline | 72% | — |
| τ²-bench Banking | 25.3% | — |
| τ²-bench Retail | 72.4% | — |
| τ²-bench Telecom | 84.9% | — |
| Cybench | 60% | — |
| DeepResearch Bench | 52.6% | — |
| OSWorld | 62.9% | — |
| GDP.pdf | — | 17.2% |
| LMArena Search | 1159 | — |
| METR Time Horizons | 67.4% | — |
Reasoning Grok 4.6 leads
Claude Sonnet 4.5: 26.9 (#125), Grok 4.6: 61.4 (#20)
| Benchmark | Claude Sonnet 4.5 | Grok 4.6 |
|---|---|---|
| ARC-AGI-2 | 13.6% | 67.1% |
| SimpleBench | 54.3% | 75.9% |
| NYT Connections (extended) | 37.3% | 80% |
| ARC-AGI-1 | 63.7% | 87.5% |
| CritPt | 1.1% | 19.7% |
| Chess Puzzles | 12% | 40% |
| EBR-Bench | 2.4% | 30.5% |
| LMArena Hard Prompts | 1462 | 1447 |
| Mystery Game Puzzles | 17% | 34% |
| DTBench | 83.2% | 97.3% |
| LMCA | 38.8% | 48.5% |
| Epoch Capabilities Index | 146.84 | 156.44 |
| Kagi LLM Benchmark | 57.9% | — |
| EnigmaEval | 6% | — |
| ForecastBench | 61.9 | — |
Math Grok 4.6 leads
Claude Sonnet 4.5: 32.3 (#216), Grok 4.6: 67.0 (#24)
| Benchmark | Claude Sonnet 4.5 | Grok 4.6 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 23.9% | 66% |
| FrontierMath Tier 4 | 2.4% | 31.7% |
| OTIS Mock AIME 2024-2025 | 77.8% | 99.2% |
| ProofBench | 19% | 51% |
| LMArena Math | 1449 | 1423 |
| Omni-MATH | 55.3% | — |
| MATH Level 5 | 97.7% | — |
| FrontierMath (Feb 2025 set) | 15.2% | — |
| FrontierMath Tier 4 (v1) | 4.2% | — |
Knowledge Grok 4.6 leads
Claude Sonnet 4.5: 48.4 (#76), Grok 4.6: 63.3 (#20)
| Benchmark | Claude Sonnet 4.5 | Grok 4.6 |
|---|---|---|
| GPQA Diamond | 82.3% | 94% |
| SimpleQA Verified | 30.7% | 49.3% |
| LMArena Expert | 1482 | 1467 |
| Humanity's Last Exam | 13.7% | — |
| MMLU-Pro | 86.9% | — |
| Vectara Hallucination Rate | 12% | — |
| GPQA (HELM) | 68.6% | — |
Multimodal Grok 4.6 leads
Claude Sonnet 4.5: 34.8 (#89), Grok 4.6: 43.6 (#23)
| Benchmark | Claude Sonnet 4.5 | Grok 4.6 |
|---|---|---|
| LMArena Document | 1450 | 1452 |
| LMArena Vision | — | 1263 |
| VPCT | 39.8% | — |
| Blueprint-Bench 2 | — | 33.2% |
| Furniture Assembly | — | 40% |
Multilingual Too close to call
Claude Sonnet 4.5: 53.4 (#69), Grok 4.6: 53.0 (#74)
| Benchmark | Claude Sonnet 4.5 | Grok 4.6 |
|---|---|---|
| LMArena Non-English | 1425 | 1420 |
| LMArena Chinese | 1459 | 1480 |
| LMArena French | 1458 | 1461 |
| LMArena German | 1427 | 1431 |
| LMArena Japanese | 1390 | 1376 |
| LMArena Korean | 1403 | 1397 |
| LMArena Russian | 1437 | 1422 |
| LMArena Spanish | 1457 | 1404 |
Instruction Following Too close to call
Claude Sonnet 4.5: 75.0 (#78), Grok 4.6: 75.4 (#63)
| Benchmark | Claude Sonnet 4.5 | Grok 4.6 |
|---|---|---|
| LMArena Instruction Following | 1459 | 1431 |
| IFEval | 85% | — |
Long Context Too close to call
Claude Sonnet 4.5: 45.2 (#46), Grok 4.6: 44.5 (#66)
| Benchmark | Claude Sonnet 4.5 | Grok 4.6 |
|---|---|---|
| LMArena Longer Query | 1476 | 1454 |
Writing & Preference Claude Sonnet 4.5 leads
Claude Sonnet 4.5: 66.5 (#34), Grok 4.6: 62.3 (#80)
| Benchmark | Claude Sonnet 4.5 | Grok 4.6 |
|---|---|---|
| LMArena Text | 1439 | 1428 |
| LMArena Creative Writing | 1442 | 1428 |
| LMArena Multi-Turn | 1465 | 1425 |
| EQ-Bench Creative Writing | 1678 | — |
| WildBench | 85.4% | — |
Frequently asked questions
Is Claude Sonnet 4.5 better than Grok 4.6?
Grok 4.6 is the stronger model overall, scoring 56.9 to 44.1 on the Noometry Index.
Which is cheaper, Claude Sonnet 4.5 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 4.5 lists at $3 and $15.
Is Claude Sonnet 4.5 or Grok 4.6 better for coding?
Grok 4.6 scores higher on coding benchmarks: 58.5 versus 47.3 in the Noometry coding category.
Which has the bigger context window?
Grok 4.6 does, with 500K tokens against 200K.
How many benchmarks do Claude Sonnet 4.5 and Grok 4.6 share?
40 benchmarks have published results for both models. Claude Sonnet 4.5 has 73 scored results on Noometry and Grok 4.6 has 49.