Muse Wonder vs Hy4 preview
Muse Wonder leads Hy4 preview on coding, 71.9 to 44.3 on the Vibe Coding Index, and Hy4 preview costs 8.9x less per output token ($2.2509 vs $20.00 per million). Updated October 21, 2026; 18 benchmarks have a result for both.
Muse Wonder or Hy4 preview: which to pick
Pick Muse Wonder for coding: it scores 71.9 on the Vibe Coding Index against 44.3 for Hy4 preview.
Pick Hy4 preview for output tokens 8.9x cheaper ($2.2509 vs $20.00 per million) and work measured by LMArena WebDev: React or BioMysteryBench (Vals AI), where it leads.
- Shared results
- 18
- Muse Wonder only
- 42
- Hy4 preview only
- 6
- Like-for-like categories
- 4 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
Muse Wonder
Muse Wonder scores 71.9 against 44.3.
3 benchmarks behind each score
Web design
Design Arena: websites
No call
Only Muse Wonder has a Design Arena: websites result, 1362.
Directional only
General
Vals Index
Muse Wonder
Muse Wonder leads, 64.9% against 53.2%.
Same board, both measured
Long documents
Listed context window
No call
Both take 1.05M tokens.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Hy4 preview
Muse Wonder $0.014, Hy4 preview $0.0019 per turn.
List prices
At a glance
- Coding index
- Muse Wonder 71.9 from 3 benchmarks, Hy4 preview 44.3 from 3 benchmarks.
- Widest category
- Coding, 12.7 points apart on average over 8 shared benchmarks.
- Widest single gap
- Terminal-Bench 4 (Vals AI): Muse Wonder 49.5% against Hy4 preview 8.1%, 55.2 points apart on a 0 to 100 scale.
- Output price
- Muse Wonder $20.00 against Hy4 preview $2.2509 per million output tokens, 8.9x apart.
- Context window
- Both take 1.05M 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 | Muse Wonder | Bars | Hy4 preview | Ties |
|---|---|---|---|---|
| Coding | 7 | 1 | ||
| Agentic work | 6 | 1 | ||
| Math | 1 | 0 | ||
| General | 2 | 0 | ||
| All shared boards | 16 | 2 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index71
| Benchmark | Muse Wonder | Hy4 preview | Leader |
|---|---|---|---|
| LMArena WebDev | 1652 | 1631 | Muse Wonder |
| Terminal-Bench 4 (Vals AI) | 49.5% | 8.1% | Muse Wonder |
| Vibe Code Bench (Vals AI) | 83.1% | 77.5% | Muse Wonder |
| Code migration (Vals AI) | 62.5% | 47.4% | Muse Wonder |
| ProgramBench (Vals AI) | Not measured | 0.0% | Only Hy4 preview measured |
| IOI (Vals AI) | 66.9% | 59.3% | Muse Wonder |
| Terminal-Bench 2.1 (Vals AI) | Not measured | 55.1% | Only Hy4 preview 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% | 65.4% | Muse Wonder |
| SciCode | 60.3% | Not measured | Only Muse Wonder measured |
| LMArena Text: coding | 1522 | Not measured | Only Muse Wonder measured |
| LMArena Text: software and IT | 1515 | Not measured | Only Muse Wonder measured |
| LMArena WebDev: React | 1641 | 1653 | Hy4 preview |
| LMArena WebDev: HTML | 1694 | 1620 | Muse Wonder |
| Neither model has a result on the other 24 boards here | |||
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index1 board, none shared
| Benchmark | Muse Wonder | Hy4 preview | Leader |
|---|---|---|---|
| Epoch Capabilities Index | 161.8 | Not measured | Only Muse Wonder measured |
| Neither model has a result on the other 1 board here | |||
Agentic workLike-for-likePublic boards, shown as published, kept out of the index61
| Benchmark | Muse Wonder | Hy4 preview | Leader |
|---|---|---|---|
| Design Arena: spreadsheets | 1227 | Not measured | Only Muse Wonder measured |
| SREBench (Vals AI) | Not measured | 2.3% | Only Hy4 preview measured |
| Finance Agent (Vals AI) | 56.6% | 55.1% | Muse Wonder |
| Tax Agent Bench (Vals AI) | 36.3% | 32.6% | Muse Wonder |
| Legal Research Bench (Vals AI) | 50.5% | 45.2% | Muse Wonder |
| Harvey's Legal Agent Benchmark (Vals AI) | 23.8% | 8.8% | Muse Wonder |
| Excel Modeling Benchmark (Vals AI) | 67.1% | 58.1% | Muse Wonder |
| BioMysteryBench (Vals AI) | 67.0% | 69.3% | Hy4 preview |
| Terminal-Bench-Science (Vals AI) | 22.9% | 1.4% | Muse Wonder |
| Vals Science Index | 39.1% | Not measured | Only Muse Wonder measured |
| CUA-bench (Vals AI) | 7.2% | 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 29 boards here | |||
ReasoningDirectional onlyPublic boards, shown as published, kept out of the index5 boards, none shared
| Benchmark | Muse Wonder | Hy4 preview | Leader |
|---|---|---|---|
| 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 |
| LMArena Text: hard prompts | 1506 | Not measured | Only Muse Wonder measured |
| LMArena Text: expert | 1513 | Not measured | Only Muse Wonder measured |
| LMArena Text: hard prompts in English | 1506 | Not measured | Only Muse Wonder measured |
| Neither model has a result on the other 17 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Muse Wonder | Hy4 preview | Leader |
|---|---|---|---|
| ProofBench | 84.0% | 75.0% | Muse Wonder |
| LMArena Text: math | 1486 | Not measured | Only Muse Wonder measured |
| LMArena Text: mathematical work | 1514 | Not measured | Only Muse Wonder measured |
| Neither model has a result on the other 10 boards here | |||
KnowledgeDirectional onlyPublic boards, shown as published, kept out of the index6 boards, none shared
| Benchmark | Muse Wonder | Hy4 preview | Leader |
|---|---|---|---|
| LegalBench (Vals AI) | Not measured | 83.8% | Only Hy4 preview measured |
| MedCode (Vals AI) | Not measured | 43.3% | Only Hy4 preview measured |
| LMArena Text: science | 1506 | Not measured | Only Muse Wonder measured |
| LMArena Text: medicine and healthcare | 1512 | Not measured | Only Muse Wonder measured |
| LMArena Text: legal and government | 1505 | Not measured | Only Muse Wonder measured |
| LMArena Text: business and finance | 1483 | Not measured | Only Muse Wonder measured |
| Neither model has a result on the other 12 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index6 boards, none shared
| Benchmark | Muse Wonder | Hy4 preview | Leader |
|---|---|---|---|
| Design Arena: websites | 1362 | Not measured | Only Muse Wonder measured |
| Design Arena: UI components | 1374 | Not measured | Only Muse Wonder measured |
| Design Arena: game dev | 1377 | Not measured | Only Muse Wonder measured |
| Design Arena: data viz | 1321 | Not measured | Only Muse Wonder measured |
| Design Arena: 3D | 1426 | Not measured | Only Muse Wonder measured |
| Design Arena: all code categories | 1371 | Not measured | Only Muse Wonder measured |
| Neither model has a result on the other 3 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | Muse Wonder | Hy4 preview | Leader |
|---|---|---|---|
| Vals Index | 64.9% | 53.2% | Muse Wonder |
| Public Benefits Bench (Vals AI) | 70.2% | 68.6% | Muse Wonder |
| MedScribe (Vals AI) | Not measured | 83.6% | Only Hy4 preview measured |
| Long context reasoning (AA-LCR) | 80.7% | Not measured | Only Muse Wonder measured |
| DTBench | 95.2% | Not measured | Only Muse Wonder measured |
| LMCA | 56.8% | Not measured | Only Muse Wonder measured |
| LMArena Text | 1502 | Not measured | Only Muse Wonder measured |
| LMArena Text: instruction following | 1491 | Not measured | Only Muse Wonder measured |
| LMArena Text: creative writing | 1491 | Not measured | Only Muse Wonder measured |
| LMArena Text: English | 1507 | Not measured | Only Muse Wonder measured |
| LMArena Text: non-English | 1492 | Not measured | Only Muse Wonder measured |
| LMArena Text: Chinese | 1537 | Not measured | Only Muse Wonder measured |
| LMArena Text: French | 1523 | Not measured | Only Muse Wonder measured |
| LMArena Text: Russian | 1522 | Not measured | Only Muse Wonder measured |
| LMArena Text: multi-turn | 1510 | Not measured | Only Muse Wonder measured |
| LMArena Text: longer queries | 1493 | Not measured | Only Muse Wonder measured |
| LMArena Text: writing and language | 1496 | Not measured | Only Muse Wonder measured |
| LMArena Text: entertainment and media | 1482 | Not measured | Only Muse Wonder measured |
| Neither model has a result on the other 28 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Muse Wonder | Hy4 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
Hy4 preview costs less
Repository review
50K input + 3K output tokens
Hy4 preview costs less
Agent loop
200K input + 10K output tokens
Hy4 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 | Meta Muse Wonder | Tencent Hy4 preview |
|---|---|---|
| Maker | Meta | Tencent |
| Weights | No public weights listed | Open weights |
| Released | October 12, 2026 | August 28, 2026 |
| Context window | 1.05M | 1.05M |
| Input $/Mtok | $4.00 | $0.7506 |
| Output $/Mtok | $20.00 | $2.2509 |
| Coding index | 71.9 | 44.3 |
| Independent measurements behind it | 3 | 3 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Muse Wonder better than Hy4 preview for coding?
Pick Muse Wonder for coding: it scores 71.9 on the Vibe Coding Index against 44.3 for Hy4 preview. Pick Hy4 preview for output tokens 8.9x cheaper ($2.2509 vs $20.00 per million) and work measured by LMArena WebDev: React or BioMysteryBench (Vals AI), where it leads.
Which is cheaper, Muse Wonder or Hy4 preview?
Muse Wonder costs $4.00 per million input tokens and $20.00 per million output tokens; Hy4 preview costs $0.7506 and $2.2509. Cost per solved task is known for neither of them.
Which has the bigger context window, Muse Wonder or Hy4 preview?
Neither: both take 1.05M tokens.
How many benchmarks is the Muse Wonder vs Hy4 preview comparison based on?
18 benchmarks have a result for both models as of October 21, 2026. Muse Wonder's coding index rests on 3 independent benchmarks and Hy4 preview's on 3. Where only one model has a result, the page says so instead of naming a winner.