Muse Spark 1.3 vs Qwen3.8 27B
Muse Spark 1.3 
Qwen3.8 27B Verdict
- Muse Spark 1.3 (max) leads on quality: 69.3 vs 59.2.
- Muse Spark 1.3 (max) is stronger in agents & tools, coding, composite, human preference, knowledge, long context, maths, multimodal, reasoning.
- Qwen3.8 27B (xhigh) is 2.1× cheaper ($0.938 vs $2.00 per 1M blended).
- Muse Spark 1.3 (max) streams 5.3× faster (223 vs 42 tokens per second).
| Metric | Muse Spark 1.3 (max) | Qwen3.8 27B (xhigh) |
|---|---|---|
| BenchLeader Index | 69.3 | 59.2 |
| Agents & tools score | 71.1 | 67.3 |
| Coding score | 66.6 | 53.2 |
| Composite score | 90.1 | 72.0 |
| Human preference score | 69.8 | – |
| Knowledge score | 76.1 | 56.4 |
| Long context score | 68.0 | 67.5 |
| Maths score | 69.2 | – |
| Multimodal score | 67.4 | 60.7 |
| Reasoning score | 83.1 | 52.5 |
| Blended price $/M | $2.00 | $0.938 |
| Output speed | 223 tok/s | 42 tok/s |
| Time to first answer | 34.8 s | 51.0 s |
| Context window | 1.0M | 262k |
| SciCode | – | 44.7% |
| LMArena Text | 1493 | – |
| LMArena Hard Prompts | 1516 | – |
| LMArena Coding | 1537 | – |
| LMArena WebDev | 1652 | – |
| LMArena Vision | 1315 | – |
| LMArena Agent | 4.2 | – |
| AA Intelligence Index | 48.2 | 33.9 |
| AA-LCR | 83.0% | 82.0% |
| MMMU-Pro | – | 76.3% |
| AA-Omniscience | 25 | -10.0 |
| GPQA Diamond (AA) | 93.5% | 90.5% |
| Humanity's Last Exam (AA) | 48.7% | 33.9% |
| SciCode (AA) | 58.8% | 46.6% |
| LiveCodeBench | – | 84.0% |
| MMLU-Pro | – | 84.3% |
| IOI | 56.6% | 39.1% |
| LegalBench | – | 82.4% |
| TaxEval | – | 70.8% |
| Terminal-Bench 2.1 (Vals) | 79.0% | 58.4% |
| SWE-bench (Vals) | – | 86.0% |
| GPQA Diamond (Vals) | – | 88.9% |
| Vals Index | 64.5 | 48.5 |
| CritPt | 24.9% | 5.4% |
| GDPval (AA) | 60.2% | 48.2% |
| τ²-Bench Banking (AA) | 50.5% | 48.0% |
| Code Migration | 47.4% | 14.2% |
| Excel Modeling Benchmark | 67.4% | 59.7% |
| Finance Agent v2 | 60.0% | 48.5% |
| Harvey's Legal Agent Benchmark | 23.8% | 11.3% |
| Legal Research Bench | 55.3% | 36.1% |
| MedCode | – | 28.7% |
| MedScribe | – | 83.8% |
| MMMU-Pro (Vals) | – | 83.9% |
| MortgageTax | – | 64.9% |
| MysteryMechanism | 36.0% | – |
| ProgramBench | – | 0.0% |
| SAGE | – | 52.4% |
| SkillsBench | – | 38.1% |
| Terminal-Bench 4.0 (Vals) | 27.8% | 4.0% |
| Terminal-Bench Science | 14.3% | 1.4% |
| Vibe Code Bench 1-100 | 20.5% | – |
| Vibe Code Bench v1.1 | 85.9% | 64.8% |
| LMArena Maths | 1496 | – |
| LMArena Creative Writing | 1453 | – |
| LMArena Instruction Following | 1483 | – |
| LMArena Multi-turn | 1493 | – |
| LMArena Longer Queries | 1506 | – |
| LMArena Document | 1468 | – |
| Surface Evolver Bench | – | 45.0% |
Data as of 2026-09-19. Best configuration of each model; every score links to its source on the model pages.
Muse Spark 1.3 vs Qwen3.8 27B: questions
- Is Muse Spark 1.3 better than Qwen3.8 27B?
- Muse Spark 1.3 (max) leads on quality: 69.3 vs 59.2. The BenchLeader Index combines every independent quality benchmark; Muse Spark 1.3 (max) is ahead overall as of 2026-09-19, but check the category scores for your use.
- Is Muse Spark 1.3 better than Qwen3.8 27B for coding?
- Muse Spark 1.3 scores higher in coding (67 vs 53 on the category index, where 50 is average).
- Is Muse Spark 1.3 better than Qwen3.8 27B for agentic tasks?
- Muse Spark 1.3 scores higher in agentic tasks (71 vs 67 on the category index, where 50 is average).
- Which is cheaper, Muse Spark 1.3 or Qwen3.8 27B?
- Qwen3.8 27B is cheaper: $0.938 against $2.00 per million tokens, blended at three input tokens per output token.
- Which is faster, Muse Spark 1.3 or Qwen3.8 27B?
- Muse Spark 1.3 streams faster: 223 against 42 output tokens per second.
- Which has the larger context window?
- Muse Spark 1.3 accepts more context: 1.0M against 262k tokens.