Model comparison
Qwen3 235B-A22B vs Step 5 Preview
Step 5 Preview is the stronger model overall, scoring 47.9 to 43.5 on the Noometry Index.
Last verified . 15 shared benchmarks.
Alibaba (Qwen)
StepFunSummary
- They share 15 benchmarks with published results for both. Qwen3 235B-A22B scores higher in 3 categories and Step 5 Preview in 5 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Step 5 Preview leads 40.0 to 15.7.
- The biggest single-benchmark swing is CritPt: 0% for Qwen3 235B-A22B and 20.9% for Step 5 Preview.
- Qwen3 235B-A22B is cheaper at $0.70 / $2.80 per million input/output tokens, against $1 / $2.70 for Step 5 Preview.
- Step 5 Preview accepts more context: 1.02M tokens versus 131K.
- Qwen3 235B-A22B has downloadable open weights; the other is API-only.
Side by side
| Qwen3 235B-A22B | Step 5 Preview | |
|---|---|---|
| Provider | Alibaba (Qwen) | StepFun |
| Noometry Index | 43.5 | 47.9 |
| Released | 2025-04 | 2026-09-16 |
| Weights | Open | Proprietary |
| Context window | 131K | 1.02M |
| Max output | 16K | 66K |
| Input $ / M tokens | $0.70 | $1 |
| Output $ / M tokens | $2.80 | $2.70 |
| Results tracked | 49 | 18 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Step 5 Preview leads
Qwen3 235B-A22B: 44.3 (#75), Step 5 Preview: 51.8 (#36)
| Benchmark | Qwen3 235B-A22B | Step 5 Preview |
|---|---|---|
| SciCode | 42.4% | 58.9% |
| LMArena Coding | 1445 | 1480 |
| Aider Polyglot | 59.6% | — |
| LMArena WebDev | — | 1564 |
| WeirdML | 41% | — |
Agentic & Tool Use Not comparable
Qwen3 235B-A22B: 33.9 (#51), Step 5 Preview: —
| Benchmark | Qwen3 235B-A22B | Step 5 Preview |
|---|---|---|
| Berkeley Function Calling Leaderboard | 52.1% | — |
| Vending-Bench 2 | -11.34 | — |
Reasoning Step 5 Preview leads
Qwen3 235B-A22B: 15.7 (#311), Step 5 Preview: 40.0 (#57)
| Benchmark | Qwen3 235B-A22B | Step 5 Preview |
|---|---|---|
| CritPt | 0% | 20.9% |
| LMArena Hard Prompts | 1433 | 1465 |
| ARC-AGI-2 | 1.3% | — |
| SimpleBench | 31% | — |
| Kagi LLM Benchmark | 69.4% | — |
| ARC-AGI-1 | 11% | — |
| Chess Puzzles | 12% | — |
| Mystery Game Puzzles | 9% | — |
| DTBench | 80.3% | — |
| LMCA | 29.3% | — |
| Epoch Capabilities Index | 143.85 | — |
| ForecastBench | 59.7 | — |
Math Qwen3 235B-A22B leads
Qwen3 235B-A22B: 50.4 (#57), Step 5 Preview: 46.1 (#72)
| Benchmark | Qwen3 235B-A22B | Step 5 Preview |
|---|---|---|
| LMArena Math | 1432 | 1470 |
| OTIS Mock AIME 2024-2025 | 86.7% | — |
| ProofBench | — | 42% |
| Omni-MATH | 71.8% | — |
| MATH Level 5 | 68.9% | — |
| FrontierMath (Feb 2025 set) | 8.5% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge Qwen3 235B-A22B leads
Qwen3 235B-A22B: 49.6 (#73), Step 5 Preview: 41.2 (#112)
| Benchmark | Qwen3 235B-A22B | Step 5 Preview |
|---|---|---|
| LMArena Expert | 1463 | 1470 |
| GPQA Diamond | 80.1% | — |
| SimpleQA Verified | 40.4% | — |
| MMLU-Pro | 84.4% | — |
| Confabulations | 15.6% | — |
| Vectara Hallucination Rate | 9.3% | — |
| GPQA (HELM) | 72.7% | — |
Multimodal Not comparable
Qwen3 235B-A22B: —, Step 5 Preview: 41.0 (#41)
| Benchmark | Qwen3 235B-A22B | Step 5 Preview |
|---|---|---|
| LMArena Vision | — | 1267 |
Multilingual Step 5 Preview leads
Qwen3 235B-A22B: 52.3 (#89), Step 5 Preview: 53.8 (#54)
| Benchmark | Qwen3 235B-A22B | Step 5 Preview |
|---|---|---|
| LMArena Non-English | 1409 | 1432 |
| LMArena Chinese | 1481 | 1519 |
| LMArena Russian | 1411 | 1436 |
| LMArena Spanish | 1430 | 1439 |
| LMArena French | 1445 | — |
| LMArena German | 1433 | — |
| LMArena Japanese | 1399 | — |
| LMArena Korean | 1391 | — |
Instruction Following Step 5 Preview leads
Qwen3 235B-A22B: 72.6 (#136), Step 5 Preview: 76.0 (#50)
| Benchmark | Qwen3 235B-A22B | Step 5 Preview |
|---|---|---|
| LMArena Instruction Following | 1408 | 1444 |
| IFEval | 83.5% | — |
Long Context Qwen3 235B-A22B leads
Qwen3 235B-A22B: 46.1 (#26), Step 5 Preview: 44.5 (#64)
| Benchmark | Qwen3 235B-A22B | Step 5 Preview |
|---|---|---|
| LMArena Longer Query | 1426 | 1455 |
| Fiction.LiveBench | 75% | — |
Writing & Preference Step 5 Preview leads
Qwen3 235B-A22B: 59.6 (#108), Step 5 Preview: 62.8 (#71)
| Benchmark | Qwen3 235B-A22B | Step 5 Preview |
|---|---|---|
| LMArena Text | 1419 | 1442 |
| LMArena Creative Writing | 1384 | 1410 |
| LMArena Multi-Turn | 1432 | 1447 |
| Short-Story Creative Writing | 83% | — |
| EQ-Bench Creative Writing | 1366 | — |
| WildBench | 86.6% | — |
Frequently asked questions
Is Qwen3 235B-A22B better than Step 5 Preview?
Step 5 Preview is the stronger model overall, scoring 47.9 to 43.5 on the Noometry Index.
Which is cheaper, Qwen3 235B-A22B or Step 5 Preview?
Qwen3 235B-A22B is cheaper. It lists at $0.70 per million input tokens and $2.80 per million output tokens; Step 5 Preview lists at $1 and $2.70.
Is Qwen3 235B-A22B or Step 5 Preview better for coding?
Step 5 Preview scores higher on coding benchmarks: 51.8 versus 44.3 in the Noometry coding category.
Which has the bigger context window?
Step 5 Preview does, with 1.02M tokens against 131K.
How many benchmarks do Qwen3 235B-A22B and Step 5 Preview share?
15 benchmarks have published results for both models. Qwen3 235B-A22B has 49 scored results on Noometry and Step 5 Preview has 18.