Muse Wonder vs Qwen4 Max
Qwen4 Max leads Muse Wonder on coding, 74.0 to 71.8 on the Vibe Coding Index, and Qwen4 Max costs 4.7x less per output token ($4.275 vs $20.00 per million). Updated November 22, 2026; 17 benchmarks have a result for both.
Muse Wonder or Qwen4 Max: which to pick
Pick Qwen4 Max for coding: it scores 74.0 on the Vibe Coding Index against 71.8 for Muse Wonder.
Pick Muse Wonder for a larger context window (1.05M against 1M tokens) and work measured by Terminal-Bench 4 (Vals AI) or Terminal-Bench 4.0 (Artificial Analysis run), where it leads.
- Shared results
- 17
- Muse Wonder only
- 65
- Qwen4 Max only
- 9
- Like-for-like categories
- 6 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 71.8.
3 and 4 benchmarks behind the scores
Intelligence
AA Intelligence Index
Muse Wonder
Muse Wonder leads, 43.6 against 42.9.
Same board, both measured
Web design
Design Arena: websites
No call
Only Muse Wonder has a Design Arena: websites result, 1374.
Directional only
General
Vals Index
Muse Wonder
Muse Wonder leads, 64.9% against 63.2%.
Same board, both measured
Long documents
Listed context window
Muse Wonder
Muse Wonder takes 1.05M tokens against 1M.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Qwen4 Max
Muse Wonder $0.014, Qwen4 Max $0.0036 per turn.
List prices
At a glance
- Coding index
- Muse Wonder 71.8 from 3 benchmarks, Qwen4 Max 74.0 from 4 benchmarks.
- Widest category
- Reasoning, 6.7 points apart on average over 2 shared benchmarks.
- Widest single gap
- LMArena WebDev: Muse Wonder 1647 against Qwen4 Max 1751, 14.8 points apart on a 0 to 100 scale.
- Output price
- Muse Wonder $20.00 against Qwen4 Max $4.275 per million output tokens, 4.7x apart.
- Context window
- Muse Wonder takes 1.05M tokens, Qwen4 Max 1M.
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 | Muse Wonder | Bars | Qwen4 Max | Ties |
|---|---|---|---|---|
| Coding | 3 | 5 | ||
| Intelligence | 1 | 0 | ||
| Agentic work | 2 | 1 | ||
| Reasoning | 1 | 1 | ||
| Knowledge | 0 | 1 | ||
| General | 1 | 1 | ||
| All shared boards | 8 | 9 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index35
| Benchmark | Muse Wonder | Qwen4 Max | Leader |
|---|---|---|---|
| LiveBench coding | Not measured | 69.8% | Only Qwen4 Max measured |
| LMArena WebDev | 1647 | 1751 | Qwen4 Max |
| Terminal-Bench 4 (Vals AI) | 49.5% | 48.5% | Muse Wonder |
| Vibe Code Bench (Vals AI) | 83.1% | 86.7% | Qwen4 Max |
| Code migration (Vals AI) | 62.5% | Not measured | Only Muse Wonder measured |
| IOI (Vals AI) | 66.9% | Not measured | Only Muse Wonder measured |
| SWE-bench Multimodal (Vals AI) | 45.6% | Not measured | Only Muse Wonder measured |
| Vibe Code Bench 1-100 (Vals AI) | 26.6% | Not measured | Only Muse Wonder measured |
| CyberBench (Vals AI) | 80.4% | Not measured | Only Muse Wonder measured |
| SciCode | 60.3% | Not measured | Only Muse Wonder measured |
| Terminal-Bench 4.0 (Artificial Analysis run) | 52.0% | 48.5% | Muse Wonder |
| FrontierCode 2.0 | 27.9% | 26.5% | Muse Wonder |
| ALE-Bench | Not measured | 1831.0 | Only Qwen4 Max measured |
| LMArena Text: coding | 1523 | Not measured | Only Muse Wonder measured |
| LMArena Text: software and IT | 1526 | Not measured | Only Muse Wonder measured |
| LMArena WebDev: React | 1655 | 1785 | Qwen4 Max |
| LMArena WebDev: HTML | 1652 | 1752 | Qwen4 Max |
| LMArena WebDev: from an image | Not measured | 1675 | Only Qwen4 Max measured |
| GameBench 2 | 77.7 | 82.4 | Qwen4 Max |
| Terminal-Bench 4.0 | 47.0% | Not measured | Only Muse Wonder measured |
| ProgramBench | 5.5% | Not measured | Only Muse Wonder measured |
| Neither model has a result on the other 31 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Muse Wonder | Qwen4 Max | Leader |
|---|---|---|---|
| AA Intelligence Index | 43.6 | 42.9 | Muse Wonder |
| Epoch Capabilities Index | 161.7 | Not measured | Only Muse Wonder measured |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index21
| Benchmark | Muse Wonder | Qwen4 Max | Leader |
|---|---|---|---|
| Design Arena: agents building mobile apps | 1239 | Not measured | Only Muse Wonder measured |
| Design Arena: HTML slides | 1129 | Not measured | Only Muse Wonder measured |
| Design Arena: PowerPoint slides | 1034 | Not measured | Only Muse Wonder measured |
| Design Arena: spreadsheets | 1230 | Not measured | Only Muse Wonder measured |
| LiveBench research | Not measured | 84.7% | Only Qwen4 Max measured |
| LiveBench browser | Not measured | 80.7% | Only Qwen4 Max measured |
| LiveBench long agentic | Not measured | 46.6% | Only Qwen4 Max measured |
| Finance Agent (Vals AI) | 56.6% | Not measured | Only Muse Wonder measured |
| Tax Agent Bench (Vals AI) | 36.3% | 40.4% | Qwen4 Max |
| Legal Research Bench (Vals AI) | 50.5% | 41.4% | Muse Wonder |
| Harvey's Legal Agent Benchmark (Vals AI) | 23.8% | 21.3% | Muse Wonder |
| Excel Modeling Benchmark (Vals AI) | 67.1% | Not measured | Only Muse Wonder measured |
| BioMysteryBench (Vals AI) | 67.0% | Not measured | Only Muse Wonder measured |
| Terminal-Bench-Science (Vals AI) | 22.9% | Not measured | Only Muse Wonder measured |
| Vals Science Index | 39.1% | Not measured | Only Muse Wonder measured |
| CUA-bench (Vals AI) | 7.2% | Not measured | Only Muse Wonder measured |
| PostTrainBench | 24.2% | Not measured | Only Muse Wonder measured |
| Vending-Bench 2 | $9,432 | Not measured | Only Muse Wonder measured |
| Neither model has a result on the other 26 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index11
| Benchmark | Muse Wonder | Qwen4 Max | Leader |
|---|---|---|---|
| LiveBench reasoning | Not measured | 92.2% | Only Qwen4 Max measured |
| MysteryMechanism (Vals AI) | 27.5% | Not measured | Only Muse Wonder measured |
| Humanity's Last Exam (Artificial Analysis run) | 50.1% | Not measured | Only Muse Wonder measured |
| HLE-Diamond (Artificial Analysis run) | 39.0% | 44.9% | Qwen4 Max |
| CritPt v2, hard subset (Artificial Analysis run) | 50.5% | 43.0% | Muse Wonder |
| LMArena Text: hard prompts | 1514 | Not measured | Only Muse Wonder measured |
| LMArena Text: expert | 1516 | Not measured | Only Muse Wonder measured |
| LMArena Text: hard prompts in English | 1519 | Not measured | Only Muse Wonder measured |
| Neither model has a result on the other 18 boards here | |||
MathDirectional onlyPublic boards, shown as published, kept out of the index4 boards, none shared
| Benchmark | Muse Wonder | Qwen4 Max | Leader |
|---|---|---|---|
| ProofBench | 84.0% | Not measured | Only Muse Wonder measured |
| LMArena Text: math | 1493 | Not measured | Only Muse Wonder measured |
| LMArena Text: mathematical work | 1509 | Not measured | Only Muse Wonder measured |
| MathArena | 51.6% | Not measured | Only Muse Wonder measured |
| Neither model has a result on the other 10 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | Muse Wonder | Qwen4 Max | Leader |
|---|---|---|---|
| AA-Omniscience | 25.1 | 27.7 | Qwen4 Max |
| LMArena Text: science | 1507 | Not measured | Only Muse Wonder measured |
| LMArena Text: medicine and healthcare | 1500 | Not measured | Only Muse Wonder measured |
| LMArena Text: legal and government | 1514 | Not measured | Only Muse Wonder measured |
| LMArena Text: business and finance | 1490 | Not measured | Only Muse Wonder measured |
| Neither model has a result on the other 16 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index8 boards, none shared
| Benchmark | Muse Wonder | Qwen4 Max | Leader |
|---|---|---|---|
| Design Arena: websites | 1374 | Not measured | Only Muse Wonder measured |
| Design Arena: UI components | 1357 | Not measured | Only Muse Wonder measured |
| Design Arena: game dev | 1394 | Not measured | Only Muse Wonder measured |
| Design Arena: data viz | 1290 | Not measured | Only Muse Wonder measured |
| Design Arena: 3D | 1420 | Not measured | Only Muse Wonder measured |
| Design Arena: all code categories | 1384 | Not measured | Only Muse Wonder measured |
| Design Arena: SVG | 1286 | Not measured | Only Muse Wonder measured |
| LiveBench design | Not measured | 79.1% | Only Qwen4 Max measured |
| Neither model has a result on the other 1 board here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index11
| Benchmark | Muse Wonder | Qwen4 Max | Leader |
|---|---|---|---|
| Vals Index | 64.9% | 63.2% | Muse Wonder |
| LiveBench data analysis | Not measured | 83.4% | Only Qwen4 Max measured |
| Public Benefits Bench (Vals AI) | 70.2% | Not measured | Only Muse Wonder measured |
| Long context reasoning (AA-LCR) | 80.7% | Not measured | Only Muse Wonder measured |
| GDPval-AA | 1770 | 1809 | Qwen4 Max |
| DTBench | 95.2% | Not measured | Only Muse Wonder measured |
| LMCA | 56.8% | Not measured | Only Muse Wonder measured |
| LMArena Text | 1503 | Not measured | Only Muse Wonder measured |
| LMArena Text: instruction following | 1492 | Not measured | Only Muse Wonder measured |
| LMArena Text: creative writing | 1488 | Not measured | Only Muse Wonder measured |
| LMArena Text: English | 1509 | Not measured | Only Muse Wonder measured |
| LMArena Text: non-English | 1492 | Not measured | Only Muse Wonder measured |
| LMArena Text: Chinese | 1542 | Not measured | Only Muse Wonder measured |
| LMArena Text: German | 1508 | Not measured | Only Muse Wonder measured |
| LMArena Text: Japanese | 1499 | Not measured | Only Muse Wonder measured |
| LMArena Text: Korean | 1465 | Not measured | Only Muse Wonder measured |
| LMArena Text: French | 1515 | Not measured | Only Muse Wonder measured |
| LMArena Text: Spanish | 1480 | Not measured | Only Muse Wonder measured |
| LMArena Text: Russian | 1511 | Not measured | Only Muse Wonder measured |
| LMArena Text: Polish | 1535 | Not measured | Only Muse Wonder measured |
| LMArena Text: multi-turn | 1512 | Not measured | Only Muse Wonder measured |
| LMArena Text: longer queries | 1499 | Not measured | Only Muse Wonder measured |
| LMArena Text: writing and language | 1494 | Not measured | Only Muse Wonder measured |
| LMArena Text: entertainment and media | 1481 | Not measured | Only Muse Wonder measured |
| EQ-Bench 4 | 1325 | Not measured | Only Muse Wonder measured |
| Neither model has a result on the other 24 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Muse Wonder | Qwen4 Max | 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
Qwen4 Max costs less
Repository review
50K input + 3K output tokens
Qwen4 Max costs less
Agent loop
200K input + 10K output tokens
Qwen4 Max 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 | Meta Muse Wonder | Qwen4 Max |
|---|---|---|
| Maker | Meta | Qwen |
| Weights | No public weights listed | No public weights listed |
| Released | October 12, 2026 | November 16, 2026 |
| Context window | 1.05M | 1M |
| Input $/Mtok | $4.00 | $1.425 |
| Output $/Mtok | $20.00 | $4.275 |
| Coding index | 71.8 | 74.0 |
| Independent measurements behind it | 3 | 4 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Muse Wonder better than Qwen4 Max for coding?
Pick Qwen4 Max for coding: it scores 74.0 on the Vibe Coding Index against 71.8 for Muse Wonder. Pick Muse Wonder for a larger context window (1.05M against 1M tokens) and work measured by Terminal-Bench 4 (Vals AI) or Terminal-Bench 4.0 (Artificial Analysis run), where it leads.
Which is cheaper, Muse Wonder or Qwen4 Max?
Muse Wonder costs $4.00 per million input tokens and $20.00 per million output tokens; Qwen4 Max costs $1.425 and $4.275. Cost per solved task is known for neither of them.
Which has the bigger context window, Muse Wonder or Qwen4 Max?
Muse Wonder, with 1.05M tokens against 1M.
How many benchmarks is the Muse Wonder vs Qwen4 Max comparison based on?
17 benchmarks have a result for both models as of November 22, 2026. Muse Wonder's coding index rests on 3 independent benchmarks and Qwen4 Max's on 4. Where only one model has a result, the page says so instead of naming a winner.