Model comparison
Claude 3.5 Sonnet vs Claude Sonnet 4
Claude Sonnet 4 is the stronger model overall, scoring 40.6 to 34.4 on the Noometry Index.
Last verified . 45 shared benchmarks.
Summary
- They share 45 benchmarks with published results for both. Claude 3.5 Sonnet scores higher in 2 categories and Claude Sonnet 4 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Claude Sonnet 4 leads 41.5 to 18.0.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 8.5% for Claude 3.5 Sonnet and 71.1% for Claude Sonnet 4.
Side by side
| Claude 3.5 Sonnet | Claude Sonnet 4 | |
|---|---|---|
| Provider | Anthropic | Anthropic |
| Noometry Index | 34.4 | 40.6 |
| Released | 2024-06-20 | 2025-05-22 |
| Weights | Proprietary | Proprietary |
| Context window | — | 200K |
| Max output | — | 64K |
| Input $ / M tokens | — | $3 |
| Output $ / M tokens | — | $15 |
| Results tracked | 60 | 58 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Sonnet 4 leads
Claude 3.5 Sonnet: 39.5 (#145), Claude Sonnet 4: 44.0 (#70)
| Benchmark | Claude 3.5 Sonnet | Claude Sonnet 4 |
|---|---|---|
| Aider Polyglot | 51.6% | 61.3% |
| GSO | 4.6% | 4.9% |
| WeirdML | 40% | 46.1% |
| LMArena Coding | 1343 | 1414 |
| SWE-bench Verified (bash only) | — | 64.9% |
| SciCode | — | 40% |
| BigCodeBench Instruct | 46.8% | — |
| LiveBench Coding | 67.1% | — |
| BigCodeBench Complete | 58.6% | — |
| CadEval | 48% | — |
| ALE-Bench | — | 655.35 |
| HumanEval+ | 81.7% | — |
| MBPP+ | 74.3% | — |
Agentic & Tool Use Claude Sonnet 4 leads
Claude 3.5 Sonnet: 32.1 (#61), Claude Sonnet 4: 38.3 (#29)
| Benchmark | Claude 3.5 Sonnet | Claude Sonnet 4 |
|---|---|---|
| TheAgentCompany | 24% | 33.1% |
| Cybench | 17.5% | 35% |
| METR Time Horizons | 45.2% | 62% |
| DeepResearch Bench | — | 46.6% |
| OSWorld | — | 43.9% |
| BALROG | 32.6% | — |
Reasoning Too close to call
Claude 3.5 Sonnet: 22.5 (#176), Claude Sonnet 4: 23.0 (#169)
| Benchmark | Claude 3.5 Sonnet | Claude Sonnet 4 |
|---|---|---|
| SimpleBench | 41.4% | 45.5% |
| EnigmaEval | 0.9% | 3.1% |
| LMArena Hard Prompts | 1305 | 1372 |
| DTBench | 67.8% | 77.1% |
| Epoch Capabilities Index | 133.55 | 141.69 |
| ForecastBench | 60.7 | 60.2 |
| ARC-AGI-2 | — | 5.9% |
| Kagi LLM Benchmark | — | 73% |
| ARC-AGI-1 | — | 40% |
| CritPt | — | 0.3% |
| LiveBench Reasoning | 56.7% | — |
| LiveBench Data Analysis | 55% | — |
| LMCA | — | 29% |
| LiveBench | 59% | — |
Math Claude Sonnet 4 leads
Claude 3.5 Sonnet: 18.0 (#282), Claude Sonnet 4: 41.5 (#71)
| Benchmark | Claude 3.5 Sonnet | Claude Sonnet 4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 8.5% | 71.1% |
| Omni-MATH | 27.6% | 60.2% |
| LMArena Math | 1306 | 1375 |
| MATH Level 5 | 56.9% | 84.4% |
| FrontierMath (Feb 2025 set) | 2.1% | 4.1% |
| FrontierMath Tier 4 (v1) | 0% | 0% |
| LiveBench Math | 52.3% | — |
Knowledge Claude Sonnet 4 leads
Claude 3.5 Sonnet: 28.5 (#236), Claude Sonnet 4: 41.6 (#99)
| Benchmark | Claude 3.5 Sonnet | Claude Sonnet 4 |
|---|---|---|
| GPQA Diamond | 55.3% | 79.2% |
| Humanity's Last Exam | 4.1% | 7.8% |
| MMLU-Pro | 77.7% | 84.3% |
| Confabulations | 19.9% | 13.2% |
| GPQA (HELM) | 56.5% | 70.6% |
| LMArena Expert | 1264 | 1371 |
| Vectara Hallucination Rate | — | 10.3% |
| MMLU | 87.3% | — |
Multimodal Too close to call
Claude 3.5 Sonnet: 26.0 (#115), Claude Sonnet 4: 25.6 (#116)
| Benchmark | Claude 3.5 Sonnet | Claude Sonnet 4 |
|---|---|---|
| LMArena Vision | 1125 | 1191 |
| GeoBench | 62% | 37% |
| VPCT | 33% | 34% |
| Video-MME | 60% | — |
| MindCube | — | 44.8% |
Multilingual Claude Sonnet 4 leads
Claude 3.5 Sonnet: 43.4 (#180), Claude Sonnet 4: 46.9 (#151)
| Benchmark | Claude 3.5 Sonnet | Claude Sonnet 4 |
|---|---|---|
| LMArena Non-English | 1283 | 1333 |
| LMArena Chinese | 1271 | 1351 |
| LMArena French | 1304 | 1364 |
| LMArena German | 1297 | 1331 |
| LMArena Japanese | 1234 | 1301 |
| LMArena Korean | 1200 | 1293 |
| LMArena Russian | 1305 | 1355 |
| LMArena Spanish | 1289 | 1356 |
Instruction Following Claude Sonnet 4 leads
Claude 3.5 Sonnet: 69.0 (#177), Claude Sonnet 4: 71.8 (#141)
| Benchmark | Claude 3.5 Sonnet | Claude Sonnet 4 |
|---|---|---|
| IFEval | 85.5% | 84% |
| LMArena Instruction Following | 1297 | 1375 |
| LiveBench Instruction Following | 69.3% | — |
Long Context Claude 3.5 Sonnet leads
Claude 3.5 Sonnet: 39.9 (#162), Claude Sonnet 4: 33.6 (#254)
| Benchmark | Claude 3.5 Sonnet | Claude Sonnet 4 |
|---|---|---|
| LMArena Longer Query | 1311 | 1398 |
| Fiction.LiveBench | — | 46.9% |
Writing & Preference Claude Sonnet 4 leads
Claude 3.5 Sonnet: 53.0 (#159), Claude Sonnet 4: 57.3 (#127)
| Benchmark | Claude 3.5 Sonnet | Claude Sonnet 4 |
|---|---|---|
| LMArena Text | 1298 | 1351 |
| LMArena Creative Writing | 1292 | 1345 |
| Short-Story Creative Writing | 80.3% | 81.4% |
| EQ-Bench Creative Writing | 1451 | 1483 |
| WildBench | 79.2% | 83.8% |
| LMArena Multi-Turn | 1326 | 1376 |
| LiveBench Language | 53.8% | — |
Frequently asked questions
Is Claude 3.5 Sonnet better than Claude Sonnet 4?
Claude Sonnet 4 is the stronger model overall, scoring 40.6 to 34.4 on the Noometry Index.
Is Claude 3.5 Sonnet or Claude Sonnet 4 better for coding?
Claude Sonnet 4 scores higher on coding benchmarks: 44.0 versus 39.5 in the Noometry coding category.
How many benchmarks do Claude 3.5 Sonnet and Claude Sonnet 4 share?
45 benchmarks have published results for both models. Claude 3.5 Sonnet has 60 scored results on Noometry and Claude Sonnet 4 has 58.