Model comparison
Claude Sonnet 4 vs Step 3
Claude Sonnet 4 and Step 3 score almost the same on the Noometry Index (40.8 vs 40.5), so choose on price, context window or the category you care about most.
Last verified . 17 shared benchmarks.
StepFunSummary
- They share 17 benchmarks with published results for both. Claude Sonnet 4 scores higher in 6 categories and Step 3 in 3 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in multimodal, where Step 3 leads 35.5 to 26.2.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 73% for Claude Sonnet 4 and 62.3% for Step 3.
- Step 3 has downloadable open weights; the other is API-only.
Side by side
| Claude Sonnet 4 | Step 3 | |
|---|---|---|
| Provider | Anthropic | StepFun |
| Noometry Index | 40.8 | 40.5 |
| Released | 2025-05-22 | — |
| Weights | Proprietary | Open |
| Context window | 200K | — |
| Max output | 64K | — |
| Input $ / M tokens | $3 | — |
| Output $ / M tokens | $15 | — |
| Results tracked | 58 | 17 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Sonnet 4 leads
Claude Sonnet 4: 43.5 (#88), Step 3: 40.1 (#147)
| Benchmark | Claude Sonnet 4 | Step 3 |
|---|---|---|
| LMArena Coding | 1414 | 1367 |
| SWE-bench Verified (bash only) | 64.9% | — |
| Aider Polyglot | 61.3% | — |
| SciCode | 40% | — |
| GSO | 4.9% | — |
| WeirdML | 46.1% | — |
| ALE-Bench | 655.35 | — |
Agentic & Tool Use Not comparable
Claude Sonnet 4: 38.5 (#31), Step 3: —
| Benchmark | Claude Sonnet 4 | Step 3 |
|---|---|---|
| TheAgentCompany | 33.1% | — |
| Cybench | 35% | — |
| DeepResearch Bench | 46.6% | — |
| OSWorld | 43.9% | — |
| METR Time Horizons | 62% | — |
Reasoning Step 3 leads
Claude Sonnet 4: 22.9 (#187), Step 3: 28.4 (#105)
| Benchmark | Claude Sonnet 4 | Step 3 |
|---|---|---|
| Kagi LLM Benchmark | 73% | 62.3% |
| LMArena Hard Prompts | 1372 | 1355 |
| ARC-AGI-2 | 5.9% | — |
| SimpleBench | 45.5% | — |
| ARC-AGI-1 | 40% | — |
| CritPt | 0.3% | — |
| EnigmaEval | 3.1% | — |
| DTBench | 77.1% | — |
| LMCA | 29% | — |
| Epoch Capabilities Index | 141.69 | — |
| ForecastBench | 60.2 | — |
Math Claude Sonnet 4 leads
Claude Sonnet 4: 43.3 (#80), Step 3: 37.6 (#148)
| Benchmark | Claude Sonnet 4 | Step 3 |
|---|---|---|
| LMArena Math | 1375 | 1366 |
| OTIS Mock AIME 2024-2025 | 71.1% | — |
| Omni-MATH | 60.2% | — |
| MATH Level 5 | 84.4% | — |
| FrontierMath (Feb 2025 set) | 4.1% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge Claude Sonnet 4 leads
Claude Sonnet 4: 41.8 (#108), Step 3: 36.8 (#164)
| Benchmark | Claude Sonnet 4 | Step 3 |
|---|---|---|
| LMArena Expert | 1372 | 1333 |
| GPQA Diamond | 79.2% | — |
| Humanity's Last Exam | 7.8% | — |
| MMLU-Pro | 84.3% | — |
| Confabulations | 13.2% | — |
| Vectara Hallucination Rate | 10.3% | — |
| GPQA (HELM) | 70.6% | — |
Multimodal Step 3 leads
Claude Sonnet 4: 26.2 (#121), Step 3: 35.5 (#86)
| Benchmark | Claude Sonnet 4 | Step 3 |
|---|---|---|
| LMArena Vision | 1191 | 1177 |
| GeoBench | 37% | — |
| VPCT | 34% | — |
| MindCube | 44.8% | — |
Multilingual Too close to call
Claude Sonnet 4: 46.7 (#156), Step 3: 46.3 (#159)
| Benchmark | Claude Sonnet 4 | Step 3 |
|---|---|---|
| LMArena Non-English | 1333 | 1327 |
| LMArena Chinese | 1350 | 1397 |
| LMArena German | 1331 | 1371 |
| LMArena Korean | 1291 | 1269 |
| LMArena Russian | 1355 | 1331 |
| LMArena Spanish | 1357 | 1371 |
| LMArena French | 1363 | — |
| LMArena Japanese | 1302 | — |
Instruction Following Claude Sonnet 4 leads
Claude Sonnet 4: 71.7 (#145), Step 3: 70.4 (#164)
| Benchmark | Claude Sonnet 4 | Step 3 |
|---|---|---|
| LMArena Instruction Following | 1376 | 1332 |
| IFEval | 84% | — |
Long Context Step 3 leads
Claude Sonnet 4: 33.7 (#259), Step 3: 40.3 (#157)
| Benchmark | Claude Sonnet 4 | Step 3 |
|---|---|---|
| LMArena Longer Query | 1398 | 1326 |
| Fiction.LiveBench | 46.9% | — |
Writing & Preference Claude Sonnet 4 leads
Claude Sonnet 4: 57.1 (#132), Step 3: 54.3 (#151)
| Benchmark | Claude Sonnet 4 | Step 3 |
|---|---|---|
| LMArena Text | 1351 | 1350 |
| LMArena Creative Writing | 1345 | 1321 |
| LMArena Multi-Turn | 1376 | 1341 |
| Short-Story Creative Writing | 81.4% | — |
| EQ-Bench Creative Writing | 1483 | — |
| WildBench | 83.8% | — |
Frequently asked questions
Is Claude Sonnet 4 better than Step 3?
Claude Sonnet 4 and Step 3 score almost the same on the Noometry Index (40.8 vs 40.5), so choose on price, context window or the category you care about most.
Is Claude Sonnet 4 or Step 3 better for coding?
Claude Sonnet 4 scores higher on coding benchmarks: 43.5 versus 40.1 in the Noometry coding category.
How many benchmarks do Claude Sonnet 4 and Step 3 share?
17 benchmarks have published results for both models. Claude Sonnet 4 has 58 scored results on Noometry and Step 3 has 17.