Model comparison
Qwen Max vs Step 3
Step 3 is the stronger model overall, scoring 40.5 to 34.7 on the Noometry Index.
Last verified . 15 shared benchmarks.
Alibaba (Qwen)
StepFunSummary
- They share 15 benchmarks with published results for both. Qwen Max scores higher in 0 categories and Step 3 in 8 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Step 3 leads 37.6 to 22.3.
- Step 3 has downloadable open weights; the other is API-only.
Side by side
| Qwen Max | Step 3 | |
|---|---|---|
| Provider | Alibaba (Qwen) | StepFun |
| Noometry Index | 34.7 | 40.5 |
| Released | 2024-04-03 | — |
| Weights | Proprietary | Open |
| Context window | 33K | — |
| Max output | 8K | — |
| Input $ / M tokens | $1.60 | — |
| Output $ / M tokens | $6.40 | — |
| Results tracked | 23 | 17 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Step 3 leads
Qwen Max: 30.7 (#292), Step 3: 40.1 (#147)
| Benchmark | Qwen Max | Step 3 |
|---|---|---|
| LMArena Coding | 1288 | 1367 |
| Aider Polyglot | 21.8% | — |
Reasoning Step 3 leads
Qwen Max: 25.1 (#151), Step 3: 28.4 (#105)
| Benchmark | Qwen Max | Step 3 |
|---|---|---|
| LMArena Hard Prompts | 1269 | 1355 |
| Kagi LLM Benchmark | — | 62.3% |
Math Step 3 leads
Qwen Max: 22.3 (#276), Step 3: 37.6 (#148)
| Benchmark | Qwen Max | Step 3 |
|---|---|---|
| LMArena Math | 1275 | 1366 |
| OTIS Mock AIME 2024-2025 | 16.1% | — |
| MATH Level 5 | 67.2% | — |
| FrontierMath (Feb 2025 set) | 1% | — |
Knowledge Step 3 leads
Qwen Max: 30.3 (#228), Step 3: 36.8 (#164)
| Benchmark | Qwen Max | Step 3 |
|---|---|---|
| LMArena Expert | 1248 | 1333 |
| GPQA Diamond | 56.1% | — |
Multimodal Not comparable
Qwen Max: —, Step 3: 35.5 (#86)
| Benchmark | Qwen Max | Step 3 |
|---|---|---|
| LMArena Vision | — | 1177 |
Multilingual Step 3 leads
Qwen Max: 41.8 (#202), Step 3: 46.3 (#159)
| Benchmark | Qwen Max | Step 3 |
|---|---|---|
| LMArena Non-English | 1263 | 1327 |
| LMArena Chinese | 1254 | 1397 |
| LMArena German | 1254 | 1371 |
| LMArena Korean | 1142 | 1269 |
| LMArena Russian | 1274 | 1331 |
| LMArena Spanish | 1290 | 1371 |
| LMArena French | 1330 | — |
| LMArena Japanese | 1205 | — |
Instruction Following Step 3 leads
Qwen Max: 66.5 (#208), Step 3: 70.4 (#164)
| Benchmark | Qwen Max | Step 3 |
|---|---|---|
| LMArena Instruction Following | 1262 | 1332 |
Long Context Too close to call
Qwen Max: 39.4 (#180), Step 3: 40.3 (#157)
| Benchmark | Qwen Max | Step 3 |
|---|---|---|
| LMArena Longer Query | 1288 | 1326 |
| Fiction.LiveBench | 66.7% | — |
Writing & Preference Step 3 leads
Qwen Max: 47.8 (#205), Step 3: 54.3 (#151)
| Benchmark | Qwen Max | Step 3 |
|---|---|---|
| LMArena Text | 1282 | 1350 |
| LMArena Creative Writing | 1248 | 1321 |
| LMArena Multi-Turn | 1277 | 1341 |
Frequently asked questions
Is Qwen Max better than Step 3?
Step 3 is the stronger model overall, scoring 40.5 to 34.7 on the Noometry Index.
Is Qwen Max or Step 3 better for coding?
Step 3 scores higher on coding benchmarks: 40.1 versus 30.7 in the Noometry coding category.
How many benchmarks do Qwen Max and Step 3 share?
15 benchmarks have published results for both models. Qwen Max has 23 scored results on Noometry and Step 3 has 17.