Model comparison
Claude 3.7 Sonnet vs Grok 4
Grok 4 is the stronger model overall, scoring 48.1 to 39.5 on the Noometry Index.
Last verified . 39 shared benchmarks.
Summary
- They share 39 benchmarks with published results for both. Claude 3.7 Sonnet scores higher in 1 category and Grok 4 in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Grok 4 leads 36.7 to 18.6.
- The biggest single-benchmark swing is ARC-AGI-1: 28.6% for Claude 3.7 Sonnet and 66.7% for Grok 4.
Side by side
| Claude 3.7 Sonnet | Grok 4 | |
|---|---|---|
| Provider | Anthropic | xAI |
| Noometry Index | 39.5 | 48.1 |
| Released | 2025-02-24 | 2025-07-09 |
| Weights | Proprietary | Proprietary |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 58 | 48 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4 leads
Claude 3.7 Sonnet: 40.6 (#136), Grok 4: 50.3 (#46)
| Benchmark | Claude 3.7 Sonnet | Grok 4 |
|---|---|---|
| Aider Polyglot | 64.9% | 79.6% |
| LMArena Coding | 1361 | 1408 |
| SWE-bench Verified | 61% | — |
| SWE-bench Verified (bash only) | 52.8% | — |
| GSO | 3.8% | — |
| WeirdML | — | 45.7% |
| LiveBench Coding | 74.5% | — |
| CadEval | 54% | — |
Agentic & Tool Use Claude 3.7 Sonnet leads
Claude 3.7 Sonnet: 34.1 (#50), Grok 4: 32.3 (#68)
| Benchmark | Claude 3.7 Sonnet | Grok 4 |
|---|---|---|
| Cybench | 20% | 43% |
| DeepResearch Bench | 43.6% | 47.3% |
| METR Time Horizons | 60% | 66.6% |
| Terminal-Bench | — | 27.2% |
| Berkeley Function Calling Leaderboard | — | 63% |
| GDPval | — | 21.1% |
| TheAgentCompany | 30.9% | — |
| OSWorld | 35.8% | — |
| BALROG | — | 43.6% |
| LMArena Search | — | 1142 |
Reasoning Grok 4 leads
Claude 3.7 Sonnet: 18.6 (#277), Grok 4: 36.7 (#65)
| Benchmark | Claude 3.7 Sonnet | Grok 4 |
|---|---|---|
| ARC-AGI-2 | 0.9% | 16% |
| SimpleBench | 46.4% | 60.5% |
| ARC-AGI-1 | 28.6% | 66.7% |
| LMArena Hard Prompts | 1333 | 1409 |
| Epoch Capabilities Index | 141.16 | 146.44 |
| ForecastBench | 61.8 | 60.9 |
| Kagi LLM Benchmark | — | 73.6% |
| Chess Puzzles | — | 28% |
| EnigmaEval | 4.2% | — |
| LiveBench Reasoning | 87.8% | — |
| LiveBench Data Analysis | 74% | — |
| LiveBench | 76.1% | — |
Math Grok 4 leads
Claude 3.7 Sonnet: 37.5 (#153), Grok 4: 48.4 (#64)
| Benchmark | Claude 3.7 Sonnet | Grok 4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 57.8% | 84% |
| Omni-MATH | 33% | 60.3% |
| LMArena Math | 1337 | 1422 |
| FrontierMath (Feb 2025 set) | 4.1% | 19.7% |
| LiveBench Math | 79% | — |
| MATH Level 5 | 91.2% | — |
| FrontierMath Tier 4 (v1) | — | 2.1% |
Knowledge Grok 4 leads
Claude 3.7 Sonnet: 39.8 (#130), Grok 4: 53.8 (#55)
| Benchmark | Claude 3.7 Sonnet | Grok 4 |
|---|---|---|
| GPQA Diamond | 79.7% | 87% |
| MMLU-Pro | 78.4% | 85.1% |
| Confabulations | 14.7% | 12.4% |
| GPQA (HELM) | 60.8% | 72.7% |
| LMArena Expert | 1321 | 1415 |
| Humanity's Last Exam | 8% | — |
Multimodal Too close to call
Claude 3.7 Sonnet: 33.7 (#95), Grok 4: 33.7 (#94)
| Benchmark | Claude 3.7 Sonnet | Grok 4 |
|---|---|---|
| LMArena Vision | 1169 | 1210 |
| GeoBench | 68% | 45% |
| VPCT | 39% | — |
| SpatialViz-Bench | 33.9% | — |
Multilingual Grok 4 leads
Claude 3.7 Sonnet: 44.1 (#179), Grok 4: 51.8 (#103)
| Benchmark | Claude 3.7 Sonnet | Grok 4 |
|---|---|---|
| LMArena Non-English | 1296 | 1403 |
| LMArena Chinese | 1299 | 1427 |
| LMArena French | 1303 | 1418 |
| LMArena German | 1301 | 1429 |
| LMArena Japanese | 1267 | 1394 |
| LMArena Korean | 1249 | 1377 |
| LMArena Russian | 1311 | 1410 |
| LMArena Spanish | 1298 | 1420 |
Instruction Following Grok 4 leads
Claude 3.7 Sonnet: 72.9 (#125), Grok 4: 79.2 (#5)
| Benchmark | Claude 3.7 Sonnet | Grok 4 |
|---|---|---|
| IFEval | 83.4% | 94.9% |
| LMArena Instruction Following | 1352 | 1387 |
| LiveBench Instruction Following | 81.3% | — |
Long Context Grok 4 leads
Claude 3.7 Sonnet: 50.3 (#10), Grok 4: 63.1 (#4)
| Benchmark | Claude 3.7 Sonnet | Grok 4 |
|---|---|---|
| Fiction.LiveBench | 83.3% | 94.4% |
| LMArena Longer Query | 1373 | 1409 |
Writing & Preference Grok 4 leads
Claude 3.7 Sonnet: 54.4 (#150), Grok 4: 58.5 (#116)
| Benchmark | Claude 3.7 Sonnet | Grok 4 |
|---|---|---|
| LMArena Text | 1314 | 1411 |
| LMArena Creative Writing | 1332 | 1397 |
| Short-Story Creative Writing | 81.1% | 76.9% |
| WildBench | 81.4% | 79.7% |
| LMArena Multi-Turn | 1339 | 1416 |
| EQ-Bench Creative Writing | 1412 | — |
| LiveBench Language | 59.9% | — |
Frequently asked questions
Is Claude 3.7 Sonnet better than Grok 4?
Grok 4 is the stronger model overall, scoring 48.1 to 39.5 on the Noometry Index.
Is Claude 3.7 Sonnet or Grok 4 better for coding?
Grok 4 scores higher on coding benchmarks: 50.3 versus 40.6 in the Noometry coding category.
How many benchmarks do Claude 3.7 Sonnet and Grok 4 share?
39 benchmarks have published results for both models. Claude 3.7 Sonnet has 58 scored results on Noometry and Grok 4 has 48.