Muse Spark 1.3 vs Qwen3.5 397B A17B
Updated October 21, 2026. 59 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.
Decision reading
Pick Muse Spark 1.3 for coding: it scores 79.7 on the Vibe Coding Index against 28.6 for Qwen3.5 397B A17B.
Pick Qwen3.5 397B A17B when output tokens cost 1.8x less.
- Shared results
- 59
- Muse Spark 1.3 only
- 33
- Qwen3.5 397B A17B only
- 6
- Like-for-like categories
- 8 of 9
Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.
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, four public coding benchmarks
Muse Spark 1.3
Muse Spark 1.3 scores 79.7 against 28.6.
2 and 1 benchmarks behind the scores
Intelligence
AA Intelligence Index
Muse Spark 1.3
Muse Spark 1.3 leads, 48.1 against 21.4.
Same board, both measured
Agentic work
AA Agentic Index
No call
Only Muse Spark 1.3 has a AA Agentic Index result, 55.5.
Directional only
Web design
Design Arena: websites
Muse Spark 1.3
Muse Spark 1.3 leads, 1356 against 1191.
Same board, both measured
Long documents
Listed context window
Muse Spark 1.3
Muse Spark 1.3 takes 1.05M tokens against 262K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Qwen3.5 397B A17B
Muse Spark 1.3 $0.0034, Qwen3.5 397B A17B $0.0016 per turn.
List prices
At a glance
- Coding index
- Muse Spark 1.3 79.7 from 2 benchmarks, Qwen3.5 397B A17B 28.6 from 1 benchmark.
- Widest category
- Agentic work, 33.7 points apart on average over 1 shared benchmark.
- Widest single gap
- FrontierMath, tiers 1 to 3: Muse Spark 1.3 74.4% against Qwen3.5 397B A17B 31.2%, 43.2 points apart on a 0 to 100 scale.
- Output price
- Muse Spark 1.3 $4.25 against Qwen3.5 397B A17B $2.34 per million output tokens, 1.8x apart.
- Context window
- Muse Spark 1.3 takes 1.05M tokens, Qwen3.5 397B A17B 262K.
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 Spark 1.3 | Bars | Qwen3.5 397B A17B | Ties |
|---|---|---|---|---|
| Coding | 7 | 0 | ||
| Intelligence | 2 | 0 | ||
| Agentic work | 1 | 0 | ||
| Reasoning | 8 | 0 | ||
| Math | 4 | 0 | ||
| Knowledge | 4 | 0 | ||
| Web design | 7 | 0 | ||
| General | 26 | 0 | ||
| All shared boards | 59 | 0 |
CodingLike-for-likeFour of these boards feed the Vibe Coding Index70
| Benchmark | Muse Spark 1.3 | Qwen3.5 397B A17B | Leader |
|---|---|---|---|
| LiveBench coding | 90.2% | Not measured | Only Muse Spark 1.3 measured |
| LMArena WebDev | 1659 | 1400 | Muse Spark 1.3 |
| AA Coding Index | 75.8 | Not measured | Only Muse Spark 1.3 measured |
| Terminal-Bench 4 (Vals AI) | 10.6% | Not measured | Only Muse Spark 1.3 measured |
| Vibe Code Bench (Vals AI) | 82.9% | Not measured | Only Muse Spark 1.3 measured |
| Code migration (Vals AI) | 27.6% | Not measured | Only Muse Spark 1.3 measured |
| ProgramBench (Vals AI) | 0.5% | Not measured | Only Muse Spark 1.3 measured |
| IOI (Vals AI) | 43.9% | Not measured | Only Muse Spark 1.3 measured |
| Terminal-Bench 2.1 (Vals AI) | 72.3% | Not measured | Only Muse Spark 1.3 measured |
| CyberBench (Vals AI) | 69.4% | Not measured | Only Muse Spark 1.3 measured |
| SciCode | 59.7% | 44.8% | Muse Spark 1.3 |
| Terminal-Bench Hard (Artificial Analysis run) | Not measured | 40.9% | Only Qwen3.5 397B A17B measured |
| FrontierSWE | 25.8% | Not measured | Only Muse Spark 1.3 measured |
| ALE-Bench | 1688.3 | 621.9 | Muse Spark 1.3 |
| LMArena Text: coding | 1510 | 1464 | Muse Spark 1.3 |
| LMArena Text: software and IT | 1511 | 1463 | Muse Spark 1.3 |
| LMArena WebDev: React | 1672 | 1388 | Muse Spark 1.3 |
| LMArena WebDev: HTML | 1653 | 1410 | Muse Spark 1.3 |
| LMArena WebDev: from an image | 1640 | Not measured | Only Muse Spark 1.3 measured |
| Neither model has a result on the other 19 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | Muse Spark 1.3 | Qwen3.5 397B A17B | Leader |
|---|---|---|---|
| AA Intelligence Index | 48.1 | 21.4 | Muse Spark 1.3 |
| Epoch Capabilities Index | 157.2 | 146.8 | Muse Spark 1.3 |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Muse Spark 1.3 | Qwen3.5 397B A17B | Leader |
|---|---|---|---|
| AA Agentic Index | 55.5 | Not measured | Only Muse Spark 1.3 measured |
| Design Arena: agents building web apps | 1297 | Not measured | Only Muse Spark 1.3 measured |
| Design Arena: agents building full-stack apps | 1295 | Not measured | Only Muse Spark 1.3 measured |
| Design Arena: agents building mobile apps | 1217 | Not measured | Only Muse Spark 1.3 measured |
| Design Arena: agents building Godot games | 1241 | Not measured | Only Muse Spark 1.3 measured |
| Design Arena: PowerPoint slides | 1244 | Not measured | Only Muse Spark 1.3 measured |
| Design Arena: spreadsheets | 1048 | Not measured | Only Muse Spark 1.3 measured |
| LiveBench research | 68.6% | Not measured | Only Muse Spark 1.3 measured |
| LiveBench browser | 80.0% | Not measured | Only Muse Spark 1.3 measured |
| LiveBench long agentic | 40.4% | Not measured | Only Muse Spark 1.3 measured |
| Finance Agent (Vals AI) | 58.9% | Not measured | Only Muse Spark 1.3 measured |
| Tax Agent Bench (Vals AI) | 36.0% | Not measured | Only Muse Spark 1.3 measured |
| Legal Research Bench (Vals AI) | 40.9% | Not measured | Only Muse Spark 1.3 measured |
| Excel Modeling Benchmark (Vals AI) | 62.7% | Not measured | Only Muse Spark 1.3 measured |
| Terminal-Bench-Science (Vals AI) | 4.3% | Not measured | Only Muse Spark 1.3 measured |
| τ²-Bench Telecom | Not measured | 95.6% | Only Qwen3.5 397B A17B measured |
| APEX-Agents | 58.6% | 24.9% | Muse Spark 1.3 |
| Neither model has a result on the other 22 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index80
| Benchmark | Muse Spark 1.3 | Qwen3.5 397B A17B | Leader |
|---|---|---|---|
| LiveBench reasoning | 89.7% | Not measured | Only Muse Spark 1.3 measured |
| GPQA Diamond (Artificial Analysis run) | 94.1% | 89.3% | Muse Spark 1.3 |
| Humanity's Last Exam (Artificial Analysis run) | 48.7% | 29.0% | Muse Spark 1.3 |
| GPQA Diamond (Epoch AI run) | Not measured | 86.4% | Only Qwen3.5 397B A17B measured |
| Chess puzzles | 37.0% | Not measured | Only Muse Spark 1.3 measured |
| Mystery game puzzles | 25.0% | 18.0% | Muse Spark 1.3 |
| LMArena Text: hard prompts | 1503 | 1448 | Muse Spark 1.3 |
| LMArena Text: expert | 1519 | 1462 | Muse Spark 1.3 |
| LMArena Text: hard prompts in English | 1500 | 1450 | Muse Spark 1.3 |
| LMArena Vision: diagrams | 1330 | 1277 | Muse Spark 1.3 |
| LMArena Vision: homework | 1296 | 1292 | Muse Spark 1.3 |
| Neither model has a result on the other 10 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index40
| Benchmark | Muse Spark 1.3 | Qwen3.5 397B A17B | Leader |
|---|---|---|---|
| OTIS Mock AIME | 99.2% | 88.9% | Muse Spark 1.3 |
| FrontierMath, tiers 1 to 3 | 74.4% | 31.2% | Muse Spark 1.3 |
| FrontierMath, tier 4 | 46.3% | Not measured | Only Muse Spark 1.3 measured |
| ProofBench | 58.0% | Not measured | Only Muse Spark 1.3 measured |
| LMArena Text: math | 1497 | 1453 | Muse Spark 1.3 |
| LMArena Text: mathematical work | 1491 | 1454 | Muse Spark 1.3 |
| Neither model has a result on the other 7 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index40
| Benchmark | Muse Spark 1.3 | Qwen3.5 397B A17B | Leader |
|---|---|---|---|
| LMArena Text: science | 1507 | 1457 | Muse Spark 1.3 |
| LMArena Text: medicine and healthcare | 1493 | 1455 | Muse Spark 1.3 |
| LMArena Text: legal and government | 1499 | 1441 | Muse Spark 1.3 |
| LMArena Text: business and finance | 1490 | 1435 | Muse Spark 1.3 |
| Neither model has a result on the other 13 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index70
| Benchmark | Muse Spark 1.3 | Qwen3.5 397B A17B | Leader |
|---|---|---|---|
| Design Arena: websites | 1356 | 1191 | Muse Spark 1.3 |
| Design Arena: UI components | 1352 | 1163 | Muse Spark 1.3 |
| Design Arena: game dev | 1329 | 1141 | Muse Spark 1.3 |
| Design Arena: data viz | 1343 | 1175 | Muse Spark 1.3 |
| Design Arena: 3D | 1359 | 1161 | Muse Spark 1.3 |
| Design Arena: all code categories | 1353 | 1180 | Muse Spark 1.3 |
| Design Arena: SVG | 1300 | 1139 | Muse Spark 1.3 |
| Design Arena: ASCII art | 1280 | Not measured | Only Muse Spark 1.3 measured |
| LiveBench design | 74.8% | Not measured | Only Muse Spark 1.3 measured |
GeneralLike-for-likePublic boards, shown as published, kept out of the index260
| Benchmark | Muse Spark 1.3 | Qwen3.5 397B A17B | Leader |
|---|---|---|---|
| LiveBench data analysis | 79.6% | Not measured | Only Muse Spark 1.3 measured |
| IFBench | Not measured | 78.8% | Only Qwen3.5 397B A17B measured |
| Long context reasoning (AA-LCR) | 83.0% | 77.3% | Muse Spark 1.3 |
| DTBench | 96.5% | 87.5% | Muse Spark 1.3 |
| LMCA | 53.9% | 37.9% | Muse Spark 1.3 |
| LMArena Text | 1490 | 1437 | Muse Spark 1.3 |
| LMArena Text: instruction following | 1479 | 1424 | Muse Spark 1.3 |
| LMArena Text: creative writing | 1462 | 1402 | Muse Spark 1.3 |
| LMArena Vision | 1308 | 1262 | Muse Spark 1.3 |
| LMArena Text: English | 1495 | 1443 | Muse Spark 1.3 |
| LMArena Text: non-English | 1481 | 1430 | Muse Spark 1.3 |
| LMArena Text: Chinese | 1532 | 1498 | Muse Spark 1.3 |
| LMArena Text: German | 1513 | 1445 | Muse Spark 1.3 |
| LMArena Text: Japanese | 1470 | 1429 | Muse Spark 1.3 |
| LMArena Text: Korean | 1490 | 1384 | Muse Spark 1.3 |
| LMArena Text: French | 1511 | 1462 | Muse Spark 1.3 |
| LMArena Text: Spanish | 1496 | 1441 | Muse Spark 1.3 |
| LMArena Text: Russian | 1491 | 1430 | Muse Spark 1.3 |
| LMArena Text: Polish | 1490 | 1447 | Muse Spark 1.3 |
| LMArena Text: multi-turn | 1487 | 1446 | Muse Spark 1.3 |
| LMArena Text: longer queries | 1488 | 1441 | Muse Spark 1.3 |
| LMArena Text: writing and language | 1470 | 1415 | Muse Spark 1.3 |
| LMArena Text: entertainment and media | 1457 | 1399 | Muse Spark 1.3 |
| LMArena Vision: English | 1298 | 1268 | Muse Spark 1.3 |
| LMArena Vision: Chinese | 1384 | 1314 | Muse Spark 1.3 |
| LMArena Vision: reading text | 1319 | 1271 | Muse Spark 1.3 |
| LMArena Vision: humor | 1293 | 1262 | Muse Spark 1.3 |
| LMArena Vision: captioning | Not measured | 1230 | Only Qwen3.5 397B A17B measured |
| LMArena Vision: recognizing things | Not measured | 1265 | Only Qwen3.5 397B A17B measured |
| LMArena Vision: creative writing | 1296 | 1260 | Muse Spark 1.3 |
| Neither model has a result on the other 16 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Muse Spark 1.3 | Qwen3.5 397B A17B | 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
Qwen3.5 397B A17B costs less
Repository review
50K input + 3K output tokens
Qwen3.5 397B A17B costs less
Agent loop
200K input + 10K output tokens
Qwen3.5 397B A17B 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 Spark 1.3 | Qwen3.5 397B A17B |
|---|---|---|
| Maker | Meta | Qwen |
| Weights | No public weights listed | Open weights |
| Released | September 2, 2026 | February 16, 2026 |
| Context window | 1.05M | 262K |
| Input $/Mtok | $1.25 | $0.39 |
| Output $/Mtok | $4.25 | $2.34 |
| Coding index | 79.7 | 28.6 |
| Independent measurements behind it | 2 | 1 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Muse Spark 1.3 better than Qwen3.5 397B A17B for coding?
Pick Muse Spark 1.3 for coding: it scores 79.7 on the Vibe Coding Index against 28.6 for Qwen3.5 397B A17B. Pick Qwen3.5 397B A17B when output tokens cost 1.8x less.
Which is cheaper, Muse Spark 1.3 or Qwen3.5 397B A17B?
Muse Spark 1.3 costs $1.25 per million input tokens and $4.25 per million output tokens; Qwen3.5 397B A17B costs $0.39 and $2.34. Cost per solved task is known for neither of them.
Which has the bigger context window, Muse Spark 1.3 or Qwen3.5 397B A17B?
Muse Spark 1.3, with 1.05M tokens against 262K.
How many benchmarks is the Muse Spark 1.3 vs Qwen3.5 397B A17B comparison based on?
59 benchmarks have a result for both models as of October 21, 2026. Muse Spark 1.3's coding index rests on 2 independent benchmarks and Qwen3.5 397B A17B's on 1. Where only one model has a result, the page says so instead of naming a winner.