Qwen4 Max vs Step 5 Preview
Qwen4 Max leads Step 5 Preview on coding, 74.0 to 57.3 on the Vibe Coding Index, and Step 5 Preview costs 1.6x less per output token ($2.70 vs $4.275 per million). Updated November 22, 2026; 25 benchmarks have a result for both.
Qwen4 Max or Step 5 Preview: which to pick
Pick Qwen4 Max for coding: it scores 74.0 on the Vibe Coding Index against 57.3 for Step 5 Preview.
Pick Step 5 Preview for output tokens 1.6x cheaper ($2.70 vs $4.275 per million).
- Shared results
- 25
- Qwen4 Max only
- 1
- Step 5 Preview only
- 51
- Like-for-like categories
- 7 of 9
Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, six public coding benchmarks
Qwen4 Max
Qwen4 Max scores 74.0 against 57.3.
4 benchmarks behind each score
Intelligence
AA Intelligence Index
Qwen4 Max
Qwen4 Max leads, 42.9 against 28.5.
Same board, both measured
General
Vals Index
Qwen4 Max
Qwen4 Max leads, 63.2% against 54.8%.
Same board, both measured
Long documents
Listed context window
No call
Both take 1M tokens.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Step 5 Preview
Qwen4 Max $0.0036, Step 5 Preview $0.0024 per turn.
List prices
At a glance
- Coding index
- Qwen4 Max 74.0 from 4 benchmarks, Step 5 Preview 57.3 from 4 benchmarks.
- Widest category
- Agentic work, 19.1 points apart on average over 6 shared benchmarks.
- Widest single gap
- LiveBench browser: Qwen4 Max 80.7% against Step 5 Preview 24.8%, 55.9 points apart on a 0 to 100 scale.
- Output price
- Qwen4 Max $4.275 against Step 5 Preview $2.70 per million output tokens, 1.6x apart.
- Context window
- Both take 1M tokens.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | Qwen4 Max | Bars | Step 5 Preview | Ties |
|---|---|---|---|---|
| Coding | 10 | 0 | ||
| Intelligence | 1 | 0 | ||
| Agentic work | 6 | 0 | ||
| Reasoning | 3 | 0 | ||
| Knowledge | 1 | 0 | ||
| Web design | 1 | 0 | ||
| General | 3 | 0 | ||
| All shared boards | 25 | 0 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index100
| Benchmark | Qwen4 Max | Step 5 Preview | Leader |
|---|---|---|---|
| LiveBench coding | 69.8% | 67.6% | Qwen4 Max |
| LMArena WebDev | 1751 | 1570 | Qwen4 Max |
| Terminal-Bench 4 (Vals AI) | 48.5% | 31.8% | Qwen4 Max |
| Vibe Code Bench (Vals AI) | 86.7% | 70.4% | Qwen4 Max |
| Code migration (Vals AI) | Not measured | 39.0% | Only Step 5 Preview measured |
| IOI (Vals AI) | Not measured | 38.1% | Only Step 5 Preview measured |
| SWE-bench Multimodal (Vals AI) | Not measured | 40.8% | Only Step 5 Preview measured |
| SciCode | Not measured | 58.9% | Only Step 5 Preview measured |
| Terminal-Bench 4.0 (Artificial Analysis run) | 48.5% | 33.3% | Qwen4 Max |
| FrontierCode 2.0 | 26.5% | 12.0% | Qwen4 Max |
| ALE-Bench | 1831.0 | Not measured | Only Qwen4 Max measured |
| LMArena Text: coding | Not measured | 1481 | Only Step 5 Preview measured |
| LMArena Text: software and IT | Not measured | 1472 | Only Step 5 Preview measured |
| LMArena WebDev: React | 1785 | 1588 | Qwen4 Max |
| LMArena WebDev: HTML | 1752 | 1542 | Qwen4 Max |
| LMArena WebDev: from an image | 1675 | 1550 | Qwen4 Max |
| GameBench 2 | 82.4 | 53.2 | Qwen4 Max |
| Neither model has a result on the other 35 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Qwen4 Max | Step 5 Preview | Leader |
|---|---|---|---|
| AA Intelligence Index | 42.9 | 28.5 | Qwen4 Max |
| Neither model has a result on the other 1 board here | |||
Agentic workLike-for-likePublic boards, shown as published, kept out of the index60
| Benchmark | Qwen4 Max | Step 5 Preview | Leader |
|---|---|---|---|
| LiveBench research | 84.7% | 75.5% | Qwen4 Max |
| LiveBench browser | 80.7% | 24.8% | Qwen4 Max |
| LiveBench long agentic | 46.6% | 19.4% | Qwen4 Max |
| SREBench (Vals AI) | Not measured | 0.4% | Only Step 5 Preview measured |
| Finance Agent (Vals AI) | Not measured | 50.7% | Only Step 5 Preview measured |
| Tax Agent Bench (Vals AI) | 40.4% | 27.3% | Qwen4 Max |
| Legal Research Bench (Vals AI) | 41.4% | 40.4% | Qwen4 Max |
| Harvey's Legal Agent Benchmark (Vals AI) | 21.3% | 12.9% | Qwen4 Max |
| Excel Modeling Benchmark (Vals AI) | Not measured | 57.1% | Only Step 5 Preview measured |
| BioMysteryBench (Vals AI) | Not measured | 67.4% | Only Step 5 Preview measured |
| Terminal-Bench-Science (Vals AI) | Not measured | 2.9% | Only Step 5 Preview measured |
| Vals Science Index | Not measured | 29.6% | Only Step 5 Preview measured |
| Neither model has a result on the other 32 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index30
| Benchmark | Qwen4 Max | Step 5 Preview | Leader |
|---|---|---|---|
| LiveBench reasoning | 92.2% | 83.8% | Qwen4 Max |
| MysteryMechanism (Vals AI) | Not measured | 18.5% | Only Step 5 Preview measured |
| Humanity's Last Exam (Artificial Analysis run) | Not measured | 46.5% | Only Step 5 Preview measured |
| HLE-Diamond (Artificial Analysis run) | 44.9% | 24.3% | Qwen4 Max |
| CritPt v2, hard subset (Artificial Analysis run) | 43.0% | 32.5% | Qwen4 Max |
| LMArena Text: hard prompts | Not measured | 1463 | Only Step 5 Preview measured |
| LMArena Text: expert | Not measured | 1479 | Only Step 5 Preview measured |
| LMArena Text: hard prompts in English | Not measured | 1464 | Only Step 5 Preview measured |
| LMArena Vision: diagrams | Not measured | 1301 | Only Step 5 Preview measured |
| LMArena Vision: homework | Not measured | 1275 | Only Step 5 Preview measured |
| Neither model has a result on the other 16 boards here | |||
MathDirectional onlyPublic boards, shown as published, kept out of the index3 boards, none shared
| Benchmark | Qwen4 Max | Step 5 Preview | Leader |
|---|---|---|---|
| ProofBench | Not measured | 42.0% | Only Step 5 Preview measured |
| LMArena Text: math | Not measured | 1466 | Only Step 5 Preview measured |
| LMArena Text: mathematical work | Not measured | 1466 | Only Step 5 Preview measured |
| Neither model has a result on the other 11 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Qwen4 Max | Step 5 Preview | Leader |
|---|---|---|---|
| AA-Omniscience | 27.7 | 16.4 | Qwen4 Max |
| LMArena Text: science | Not measured | 1463 | Only Step 5 Preview measured |
| LMArena Text: medicine and healthcare | Not measured | 1443 | Only Step 5 Preview measured |
| LMArena Text: legal and government | Not measured | 1439 | Only Step 5 Preview measured |
| LMArena Text: business and finance | Not measured | 1445 | Only Step 5 Preview measured |
| Neither model has a result on the other 16 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Qwen4 Max | Step 5 Preview | Leader |
|---|---|---|---|
| LiveBench design | 79.1% | 70.8% | Qwen4 Max |
| Neither model has a result on the other 8 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index30
| Benchmark | Qwen4 Max | Step 5 Preview | Leader |
|---|---|---|---|
| Vals Index | 63.2% | 54.8% | Qwen4 Max |
| LiveBench data analysis | 83.4% | 63.5% | Qwen4 Max |
| Public Benefits Bench (Vals AI) | Not measured | 58.1% | Only Step 5 Preview measured |
| Long context reasoning (AA-LCR) | Not measured | 88.3% | Only Step 5 Preview measured |
| GDPval-AA | 1809 | 1591 | Qwen4 Max |
| LMArena Text | Not measured | 1444 | Only Step 5 Preview measured |
| LMArena Text: instruction following | Not measured | 1446 | Only Step 5 Preview measured |
| LMArena Text: creative writing | Not measured | 1416 | Only Step 5 Preview measured |
| LMArena Vision | Not measured | 1265 | Only Step 5 Preview measured |
| LMArena Text: English | Not measured | 1456 | Only Step 5 Preview measured |
| LMArena Text: non-English | Not measured | 1430 | Only Step 5 Preview measured |
| LMArena Text: Chinese | Not measured | 1495 | Only Step 5 Preview measured |
| LMArena Text: German | Not measured | 1453 | Only Step 5 Preview measured |
| LMArena Text: Japanese | Not measured | 1398 | Only Step 5 Preview measured |
| LMArena Text: Korean | Not measured | 1402 | Only Step 5 Preview measured |
| LMArena Text: French | Not measured | 1444 | Only Step 5 Preview measured |
| LMArena Text: Spanish | Not measured | 1445 | Only Step 5 Preview measured |
| LMArena Text: Russian | Not measured | 1432 | Only Step 5 Preview measured |
| LMArena Text: Polish | Not measured | 1406 | Only Step 5 Preview measured |
| LMArena Text: multi-turn | Not measured | 1439 | Only Step 5 Preview measured |
| LMArena Text: longer queries | Not measured | 1454 | Only Step 5 Preview measured |
| LMArena Text: writing and language | Not measured | 1431 | Only Step 5 Preview measured |
| LMArena Text: entertainment and media | Not measured | 1414 | Only Step 5 Preview measured |
| LMArena Vision: English | Not measured | 1275 | Only Step 5 Preview measured |
| LMArena Vision: Chinese | Not measured | 1318 | Only Step 5 Preview measured |
| LMArena Vision: reading text | Not measured | 1279 | Only Step 5 Preview measured |
| LMArena Vision: humor | Not measured | 1215 | Only Step 5 Preview measured |
| LMArena Vision: creative writing | Not measured | 1250 | Only Step 5 Preview measured |
| Neither model has a result on the other 21 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Qwen4 Max | Step 5 Preview | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
Step 5 Preview costs less
Repository review
50K input + 3K output tokens
Step 5 Preview costs less
Agent loop
200K input + 10K output tokens
Step 5 Preview costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | Qwen4 Max | StepFun Step 5 Preview |
|---|---|---|
| Maker | Qwen | StepFun |
| Weights | No public weights listed | Open weights |
| Released | November 16, 2026 | October 8, 2026 |
| Context window | 1M | 1M |
| Input $/Mtok | $1.425 | $1.00 |
| Output $/Mtok | $4.275 | $2.70 |
| Coding index | 74.0 | 57.3 |
| Independent measurements behind it | 4 | 4 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Qwen4 Max better than Step 5 Preview for coding?
Pick Qwen4 Max for coding: it scores 74.0 on the Vibe Coding Index against 57.3 for Step 5 Preview. Pick Step 5 Preview for output tokens 1.6x cheaper ($2.70 vs $4.275 per million).
Which is cheaper, Qwen4 Max or Step 5 Preview?
Qwen4 Max costs $1.425 per million input tokens and $4.275 per million output tokens; Step 5 Preview costs $1.00 and $2.70. Cost per solved task is known for neither of them.
Which has the bigger context window, Qwen4 Max or Step 5 Preview?
Neither: both take 1M tokens.
How many benchmarks is the Qwen4 Max vs Step 5 Preview comparison based on?
25 benchmarks have a result for both models as of November 22, 2026. Qwen4 Max's coding index rests on 4 independent benchmarks and Step 5 Preview's on 4. Where only one model has a result, the page says so instead of naming a winner.