Mercury 2.5 vs Muse Wonder
Muse Wonder leads Mercury 2.5 on coding, 71.9 to 1.6 on the Vibe Coding Index, and Mercury 2.5 costs 133.3x less per output token ($0.15 vs $20.00 per million). Updated October 22, 2026; 15 benchmarks have a result for both.
Mercury 2.5 or Muse Wonder: which to pick
Pick Muse Wonder for coding: it scores 71.9 on the Vibe Coding Index against 1.6 for Mercury 2.5.
Pick Mercury 2.5 for output tokens 133.3x cheaper ($0.15 vs $20.00 per million).
- Shared results
- 15
- Mercury 2.5 only
- 7
- Muse Wonder only
- 45
- Like-for-like categories
- 5 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 1.6.
2 and 3 benchmarks behind the scores
Intelligence
AA Intelligence Index
No call
Only Mercury 2.5 has a AA Intelligence Index result, 12.3.
Directional only
Web design
Design Arena: websites
No call
Only Muse Wonder has a Design Arena: websites result, 1363.
Directional only
General
Vals Index
No call
Only Muse Wonder has a Vals Index result, 64.9%.
Directional only
Long documents
Listed context window
Muse Wonder
Muse Wonder takes 1.05M tokens against 260K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Mercury 2.5
Mercury 2.5 $0.0001, Muse Wonder $0.014 per turn.
List prices
At a glance
- Coding index
- Mercury 2.5 1.6 from 2 benchmarks, Muse Wonder 71.9 from 3 benchmarks.
- Widest category
- Math, 81.0 points apart on average over 1 shared benchmark.
- Widest single gap
- ProofBench: Mercury 2.5 3.0% against Muse Wonder 84.0%, 81.0 points apart on a 0 to 100 scale.
- Output price
- Mercury 2.5 $0.15 against Muse Wonder $20.00 per million output tokens, 133.3x apart.
- Context window
- Mercury 2.5 takes 260K tokens, Muse Wonder 1.05M.
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 | Mercury 2.5 | Bars | Muse Wonder | Ties |
|---|---|---|---|---|
| Coding | 0 | 6 | ||
| Agentic work | 0 | 5 | ||
| Reasoning | 0 | 1 | ||
| Math | 0 | 1 | ||
| General | 0 | 2 | ||
| All shared boards | 0 | 15 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index06
| Benchmark | Mercury 2.5 | Muse Wonder | Leader |
|---|---|---|---|
| LMArena WebDev | Not measured | 1652 | Only Muse Wonder measured |
| Terminal-Bench 4 (Vals AI) | 0.0% | 49.5% | Muse Wonder |
| Vibe Code Bench (Vals AI) | 3.7% | 83.1% | Muse Wonder |
| Code migration (Vals AI) | 4.5% | 62.5% | Muse Wonder |
| IOI (Vals AI) | 2.4% | 66.9% | Muse Wonder |
| Terminal-Bench 2.1 (Vals AI) | 34.5% | Not measured | Only Mercury 2.5 measured |
| SWE-bench Multimodal (Vals AI) | Not measured | 45.6% | Only Muse Wonder measured |
| Vibe Code Bench 1-100 (Vals AI) | Not measured | 26.6% | Only Muse Wonder measured |
| SkillsBench (Vals AI) | 18.1% | Not measured | Only Mercury 2.5 measured |
| CyberBench (Vals AI) | 45.5% | 80.4% | Muse Wonder |
| SciCode | 38.5% | 60.3% | Muse Wonder |
| ALE-Bench | 301.6 | Not measured | Only Mercury 2.5 measured |
| LMArena Text: coding | Not measured | 1522 | Only Muse Wonder measured |
| LMArena Text: software and IT | Not measured | 1515 | Only Muse Wonder measured |
| LMArena WebDev: React | Not measured | 1641 | Only Muse Wonder measured |
| LMArena WebDev: HTML | Not measured | 1694 | Only Muse Wonder measured |
| Neither model has a result on the other 23 boards here | |||
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | Mercury 2.5 | Muse Wonder | Leader |
|---|---|---|---|
| AA Intelligence Index | 12.3 | Not measured | Only Mercury 2.5 measured |
| Epoch Capabilities Index | Not measured | 161.8 | Only Muse Wonder measured |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index05
| Benchmark | Mercury 2.5 | Muse Wonder | Leader |
|---|---|---|---|
| Design Arena: spreadsheets | Not measured | 1260 | Only Muse Wonder measured |
| Finance Agent (Vals AI) | 18.9% | 56.6% | Muse Wonder |
| Tax Agent Bench (Vals AI) | 2.5% | 36.3% | Muse Wonder |
| Legal Research Bench (Vals AI) | 4.3% | 50.5% | Muse Wonder |
| Harvey's Legal Agent Benchmark (Vals AI) | Not measured | 23.8% | Only Muse Wonder measured |
| Excel Modeling Benchmark (Vals AI) | 8.8% | 67.1% | Muse Wonder |
| BioMysteryBench (Vals AI) | Not measured | 67.0% | Only Muse Wonder measured |
| Terminal-Bench-Science (Vals AI) | 0.0% | 22.9% | Muse Wonder |
| Vals Science Index | Not measured | 39.1% | Only Muse Wonder measured |
| CUA-bench (Vals AI) | Not measured | 7.2% | Only Muse Wonder measured |
| Vending-Bench 2 | Not measured | $9,432 | Only Muse Wonder measured |
| Neither model has a result on the other 30 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | Mercury 2.5 | Muse Wonder | Leader |
|---|---|---|---|
| MysteryMechanism (Vals AI) | Not measured | 27.5% | Only Muse Wonder measured |
| Humanity's Last Exam (Artificial Analysis run) | 11.8% | 50.1% | Muse Wonder |
| LMArena Text: hard prompts | Not measured | 1506 | Only Muse Wonder measured |
| LMArena Text: expert | Not measured | 1513 | Only Muse Wonder measured |
| LMArena Text: hard prompts in English | Not measured | 1506 | 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 index01
| Benchmark | Mercury 2.5 | Muse Wonder | Leader |
|---|---|---|---|
| ProofBench | 3.0% | 84.0% | Muse Wonder |
| LMArena Text: math | Not measured | 1486 | Only Muse Wonder measured |
| LMArena Text: mathematical work | Not measured | 1514 | 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 | Mercury 2.5 | Muse Wonder | Leader |
|---|---|---|---|
| LegalBench (Vals AI) | 83.1% | Not measured | Only Mercury 2.5 measured |
| MedCode (Vals AI) | 31.3% | Not measured | Only Mercury 2.5 measured |
| LMArena Text: science | Not measured | 1506 | Only Muse Wonder measured |
| LMArena Text: medicine and healthcare | Not measured | 1512 | Only Muse Wonder measured |
| LMArena Text: legal and government | Not measured | 1505 | Only Muse Wonder measured |
| LMArena Text: business and finance | Not measured | 1483 | 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 | Mercury 2.5 | Muse Wonder | Leader |
|---|---|---|---|
| Design Arena: websites | Not measured | 1363 | Only Muse Wonder measured |
| Design Arena: UI components | Not measured | 1380 | Only Muse Wonder measured |
| Design Arena: game dev | Not measured | 1378 | Only Muse Wonder measured |
| Design Arena: data viz | Not measured | 1312 | Only Muse Wonder measured |
| Design Arena: 3D | Not measured | 1428 | Only Muse Wonder measured |
| Design Arena: all code categories | Not measured | 1371 | 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 index02
| Benchmark | Mercury 2.5 | Muse Wonder | Leader |
|---|---|---|---|
| Vals Index | Not measured | 64.9% | Only Muse Wonder measured |
| Public Benefits Bench (Vals AI) | 45.3% | 70.2% | Muse Wonder |
| MedScribe (Vals AI) | 55.1% | Not measured | Only Mercury 2.5 measured |
| Long context reasoning (AA-LCR) | 71.7% | 80.7% | Muse Wonder |
| DTBench | Not measured | 95.2% | Only Muse Wonder measured |
| LMCA | Not measured | 56.8% | Only Muse Wonder measured |
| LMArena Text | Not measured | 1502 | Only Muse Wonder measured |
| LMArena Text: instruction following | Not measured | 1491 | Only Muse Wonder measured |
| LMArena Text: creative writing | Not measured | 1491 | Only Muse Wonder measured |
| LMArena Text: English | Not measured | 1507 | Only Muse Wonder measured |
| LMArena Text: non-English | Not measured | 1492 | Only Muse Wonder measured |
| LMArena Text: Chinese | Not measured | 1537 | Only Muse Wonder measured |
| LMArena Text: French | Not measured | 1523 | Only Muse Wonder measured |
| LMArena Text: Russian | Not measured | 1522 | Only Muse Wonder measured |
| LMArena Text: multi-turn | Not measured | 1510 | Only Muse Wonder measured |
| LMArena Text: longer queries | Not measured | 1493 | Only Muse Wonder measured |
| LMArena Text: writing and language | Not measured | 1496 | Only Muse Wonder measured |
| LMArena Text: entertainment and media | Not measured | 1482 | 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 | Mercury 2.5 | Muse Wonder | 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
Mercury 2.5 costs less
Repository review
50K input + 3K output tokens
Mercury 2.5 costs less
Agent loop
200K input + 10K output tokens
Mercury 2.5 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 | Inception Mercury 2.5 | Meta Muse Wonder |
|---|---|---|
| Maker | Inception | Meta |
| Weights | No public weights listed | No public weights listed |
| Released | September 8, 2026 | October 12, 2026 |
| Context window | 260K | 1.05M |
| Input $/Mtok | $0.04 | $4.00 |
| Output $/Mtok | $0.15 | $20.00 |
| Coding index | 1.6 | 71.9 |
| Independent measurements behind it | 2 | 3 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Mercury 2.5 better than Muse Wonder for coding?
Pick Muse Wonder for coding: it scores 71.9 on the Vibe Coding Index against 1.6 for Mercury 2.5. Pick Mercury 2.5 for output tokens 133.3x cheaper ($0.15 vs $20.00 per million).
Which is cheaper, Mercury 2.5 or Muse Wonder?
Mercury 2.5 costs $0.04 per million input tokens and $0.15 per million output tokens; Muse Wonder costs $4.00 and $20.00. Cost per solved task is known for neither of them.
Which has the bigger context window, Mercury 2.5 or Muse Wonder?
Muse Wonder, with 1.05M tokens against 260K.
How many benchmarks is the Mercury 2.5 vs Muse Wonder comparison based on?
15 benchmarks have a result for both models as of October 22, 2026. Mercury 2.5's coding index rests on 2 independent benchmarks and Muse Wonder's on 3. Where only one model has a result, the page says so instead of naming a winner.