Muse Spark 1.3 vs Qwen3.8 Max
Muse Spark 1.3 
Qwen3.8 Max Verdict
- Muse Spark 1.3 (max) leads on quality: 67.7 vs 64.2.
- Muse Spark 1.3 (max) is stronger in agents & tools, composite, human preference, knowledge, long context, reasoning.
- Qwen3.8 Max (max) is stronger in maths, multimodal.
- Muse Spark 1.3 (max) is 1.5× cheaper ($2.00 vs $3.00 per 1M blended).
- Muse Spark 1.3 (max) streams 4.8× faster (175 vs 37 tokens per second).
| Metric | Muse Spark 1.3 (max) | Qwen3.8 Max (max) |
|---|---|---|
| BenchLeader Index | 67.7 | 64.2 |
| Agents & tools score | 70.3 | 58.2 |
| Coding score | 64.6 | 64.6 |
| Composite score | 84.1 | 70.7 |
| Human preference score | 69.4 | 68.0 |
| Knowledge score | 72.9 | 62.7 |
| Long context score | 66.8 | 65.4 |
| Maths score | 63.6 | 65.1 |
| Multimodal score | 65.9 | 66.2 |
| Reasoning score | 78.1 | 72.2 |
| Blended price $/M | $2.00 | $3.00 |
| Output speed | 175 tok/s | 37 tok/s |
| Time to first answer | 70.5 s | 58.9 s |
| Context window | 1.0M | 1M |
| FrontierMath Tiers 1–3 | 74.0% | – |
| FrontierMath Tier 4 | 46.3% | – |
| Terminal-Bench | – | 27.0% |
| SciCode | 58.8% | 53.2% |
| APEX-Agents | – | 63.3% |
| ProofBench | 58.0% | 58.0% |
| Epoch Capabilities Index | – | 156.4 |
| LMArena Text | 1494 | 1483 |
| LMArena Hard Prompts | 1517 | 1504 |
| LMArena Coding | 1538 | 1524 |
| LMArena WebDev | 1657 | 1672 |
| LMArena Vision | 1309 | 1314 |
| LMArena Agent | 4 | 2.3 |
| LiveBench | – | 78.5% |
| LiveBench Reasoning | – | 88.2% |
| LiveBench Coding | – | 72.9% |
| LiveBench Agentic Coding | – | 64.7% |
| LiveBench Mathematics | – | 91.3% |
| LiveBench Data Analysis | – | 78.4% |
| LiveBench Language | – | 79.7% |
| LiveBench Instruction Following | – | 74.1% |
| AA Intelligence Index v4.3.2 | 48.1 | 45.4 |
| AA-LCR | 83.0% | 80.3% |
| MMMU-Pro | – | 82.8% |
| AA-Omniscience | 25 | 12.0 |
| GPQA Diamond (AA) | 93.5% | 92.8% |
| Humanity's Last Exam (AA) | 48.7% | 43.1% |
| SciCode (AA) | 58.8% | 53.2% |
| LiveCodeBench | – | 87.8% |
| MMLU-Pro | – | 88.6% |
| IOI | 56.6% | 68.9% |
| LegalBench | – | 83.6% |
| CorpFin | – | 65.8% |
| TaxEval | – | 75.5% |
| Terminal-Bench 2.1 (Vals) | 79.0% | 67.4% |
| SWE-bench (Vals) | – | 85.6% |
| GPQA Diamond (Vals) | – | 93.7% |
| Vals Index | 58.2 | 48.3 |
| CritPt | 24.9% | 20.0% |
| GDPval-AA v2.1 | 59.0% | 58.6% |
| τ³-Banking (AA) | 50.5% | 51.3% |
| ITBench SRE (AA) | 33.2% | 40.3% |
| Analyst Agent (AA) | – | 45.0% |
| APEX-Agents (AA) | – | 42.4% |
| Code Migration | 47.4% | 24.0% |
| CUA-bench | 5.8% | – |
| CyberBench | 72.7% | 28.6% |
| Excel Modeling Benchmark | 67.4% | 60.1% |
| Finance Agent v2 | 60.0% | 50.6% |
| Harvey's Legal Agent Benchmark | 23.8% | 10.4% |
| Legal Research Bench | 55.3% | 47.6% |
| MedCode | – | 40.7% |
| MedScribe | – | 85.0% |
| MMMU-Pro (Vals) | – | 88.0% |
| MortgageTax | – | 64.0% |
| MysteryMechanism | 36.0% | 23.9% |
| ProgramBench | – | 0.0% |
| Public Benefits Bench | – | 67.1% |
| SAGE | – | 51.3% |
| SkillsBench | – | 42.0% |
| Tax Agent Bench | 72.4% | 66.0% |
| Terminal-Bench 4.0 (Vals) | 24.8% | 34.3% |
| Terminal-Bench Science | 10.0% | 1.4% |
| Vals Multimodal Index | – | 65.4% |
| Vibe Code Bench 1-100 | 20.5% | 12.8% |
| Vibe Code Bench v1.1 | 85.9% | 64.7% |
| LMArena Maths | 1502 | 1497 |
| LMArena Creative Writing | 1456 | 1470 |
| LMArena Instruction Following | 1483 | 1474 |
| LMArena Multi-turn | 1487 | 1492 |
| LMArena Longer Queries | 1497 | 1492 |
| LMArena Document | 1468 | – |
| Chess Puzzles | 38.0% | – |
| Mystery Game Puzzles | 25.0% | – |
| CursorBench | 41.6% | – |
| Terminal-Bench 4.0 (AA) | 33.3% | 38.9% |
| Terminal-Bench 2.1 (AA) | 84.3% | 88.8% |
| AutomationBench | 57.9% | 56.2% |
| GDP.pdf | 26.6% | 22.8% |
| MLCR | 43.3% | 20.0% |
| Harvey LAB | 8.9% | – |
| EnterpriseOps-Gym | – | 47.6% |
| AA-Omniscience: accuracy | 43.6% | 31.9% |
| AA-Omniscience: non-hallucination | 67.1% | 71.2% |
| AA-Briefcase v1.1 | 1581 | 1617 |
Data as of 2026-10-11. Best configuration of each model; every score links to its source on the model pages.
Muse Spark 1.3 vs Qwen3.8 Max: questions
- Is Muse Spark 1.3 better than Qwen3.8 Max?
- Muse Spark 1.3 (max) leads on quality: 67.7 vs 64.2. The BenchLeader Index combines every independent quality benchmark; Muse Spark 1.3 (max) is ahead overall as of 2026-10-11, but check the category scores for your use.
- Is Muse Spark 1.3 better than Qwen3.8 Max for coding?
- They are level in coding (65 each).
- Is Muse Spark 1.3 better than Qwen3.8 Max for agentic tasks?
- Muse Spark 1.3 scores higher in agentic tasks (70 vs 58 on the category index, where 50 is average).
- Which is cheaper, Muse Spark 1.3 or Qwen3.8 Max?
- Muse Spark 1.3 is cheaper: $2.00 against $3.00 per million tokens, blended at three input tokens per output token.
- Which is faster, Muse Spark 1.3 or Qwen3.8 Max?
- Muse Spark 1.3 streams faster: 175 against 37 output tokens per second.
- Which has the larger context window?
- Muse Spark 1.3 accepts more context: 1.0M against 1M tokens.