Model comparison
Claude 3 Sonnet vs Step 3
Step 3 is the stronger model overall, scoring 40.5 to 29.0 on the Noometry Index.
Last verified . 16 shared benchmarks.
StepFunSummary
- They share 16 benchmarks with published results for both. Claude 3 Sonnet scores higher in 0 categories and Step 3 in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Step 3 leads 37.6 to 10.7.
- Step 3 has downloadable open weights; the other is API-only.
Side by side
| Claude 3 Sonnet | Step 3 | |
|---|---|---|
| Provider | Anthropic | StepFun |
| Noometry Index | 29.0 | 40.5 |
| Released | 2024-02-29 | — |
| Weights | Proprietary | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 30 | 17 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Step 3 leads
Claude 3 Sonnet: 29.6 (#302), Step 3: 40.1 (#147)
| Benchmark | Claude 3 Sonnet | Step 3 |
|---|---|---|
| LMArena Coding | 1223 | 1367 |
| WeirdML | 10.2% | — |
| BigCodeBench Instruct | 42.7% | — |
| BigCodeBench Complete | 53.8% | — |
| HumanEval+ | 64% | — |
| MBPP+ | 69.3% | — |
Reasoning Step 3 leads
Claude 3 Sonnet: 20.5 (#237), Step 3: 28.4 (#105)
| Benchmark | Claude 3 Sonnet | Step 3 |
|---|---|---|
| LMArena Hard Prompts | 1197 | 1355 |
| Kagi LLM Benchmark | — | 62.3% |
| DTBench | 53.6% | — |
| Epoch Capabilities Index | 120.7 | — |
| WinoGrande | 75.1% | — |
Math Step 3 leads
Claude 3 Sonnet: 10.7 (#310), Step 3: 37.6 (#148)
| Benchmark | Claude 3 Sonnet | Step 3 |
|---|---|---|
| LMArena Math | 1213 | 1366 |
| OTIS Mock AIME 2024-2025 | 2.5% | — |
| MATH Level 5 | 18.2% | — |
Knowledge Step 3 leads
Claude 3 Sonnet: 21.1 (#276), Step 3: 36.8 (#164)
| Benchmark | Claude 3 Sonnet | Step 3 |
|---|---|---|
| LMArena Expert | 1173 | 1333 |
| GPQA Diamond | 40.6% | — |
| MMLU | 75.9% | — |
Multimodal Step 3 leads
Claude 3 Sonnet: 25.2 (#125), Step 3: 35.5 (#86)
| Benchmark | Claude 3 Sonnet | Step 3 |
|---|---|---|
| LMArena Vision | 984 | 1177 |
Multilingual Step 3 leads
Claude 3 Sonnet: 37.8 (#234), Step 3: 46.3 (#159)
| Benchmark | Claude 3 Sonnet | Step 3 |
|---|---|---|
| LMArena Non-English | 1205 | 1327 |
| LMArena Chinese | 1189 | 1397 |
| LMArena German | 1204 | 1371 |
| LMArena Korean | 1128 | 1269 |
| LMArena Russian | 1227 | 1331 |
| LMArena Spanish | 1204 | 1371 |
| LMArena French | 1229 | — |
| LMArena Japanese | 1131 | — |
Instruction Following Step 3 leads
Claude 3 Sonnet: 62.8 (#235), Step 3: 70.4 (#164)
| Benchmark | Claude 3 Sonnet | Step 3 |
|---|---|---|
| LMArena Instruction Following | 1199 | 1332 |
Long Context Step 3 leads
Claude 3 Sonnet: 36.7 (#228), Step 3: 40.3 (#157)
| Benchmark | Claude 3 Sonnet | Step 3 |
|---|---|---|
| LMArena Longer Query | 1211 | 1326 |
Writing & Preference Step 3 leads
Claude 3 Sonnet: 42.1 (#238), Step 3: 54.3 (#151)
| Benchmark | Claude 3 Sonnet | Step 3 |
|---|---|---|
| LMArena Text | 1218 | 1350 |
| LMArena Creative Writing | 1186 | 1321 |
| LMArena Multi-Turn | 1227 | 1341 |
Frequently asked questions
Is Claude 3 Sonnet better than Step 3?
Step 3 is the stronger model overall, scoring 40.5 to 29.0 on the Noometry Index.
Is Claude 3 Sonnet or Step 3 better for coding?
Step 3 scores higher on coding benchmarks: 40.1 versus 29.6 in the Noometry coding category.
How many benchmarks do Claude 3 Sonnet and Step 3 share?
16 benchmarks have published results for both models. Claude 3 Sonnet has 30 scored results on Noometry and Step 3 has 17.