Comparing 2 models
Muse Spark 1.3 vs Qwen3.8 27B
Benchmarks through
Verdict
Qwen3.8 27B places higher on 4 of the 4 tasks where both are ranked. Qwen3.8 27B costs 31.3x less per input token. Muse Spark 1.3 accepts 1.0x more context. Only Qwen3.8 27B has open weights for self-hosting.
- Choose Qwen3.8 27B for Value LLM (#1), Fast Summaries (#2), Structured Data Extraction (#5), you need to self-host.
Built from the published task rankings below; ranks come from each task's own weighted benchmarks, so a win on one task says nothing about another.
Measured head to head
Benchmarks both models were run on
| Evaluation | Muse Spark 1.3 | Qwen3.8 27B | Source |
|---|---|---|---|
| LiveBench Agentic Coding | 64.1% | 61.4% | source · 2026-10-18 |
| LiveBench Coding | 81.1% | 75.7% | source · 2026-10-18 |
| LiveBench Data Analysis | 79.6% | 76.6% | source · 2026-10-18 |
| LiveBench Instruction Following | 78% | 72.7% | source · 2026-10-18 |
| LiveBench Language | 82.8% | 74.3% | source · 2026-10-18 |
| LiveBench Math | 96.0% | 86.2% | source · 2026-10-18 |
| LiveBench Reasoning | 89.7% | 80.0% | source · 2026-10-18 |
| LMArena Agent | 3.4 | -3.8 | source · 2026-10-20 |
| LMArena Agent Steerability | 5 | -3.9 | source · 2026-10-20 |
| LMArena Agent Task Outcome | 5.3 | -1.5 | source · 2026-10-20 |
| LMArena Agent Tool Hallucination | 0.6 | -0.1 | source · 2026-10-20 |
| LMArena Creative Writing | 1462 Elo | 1383 Elo | source · 2026-10-20 |
| LMArena Instruction Following | 1479 Elo | 1436 Elo | source · 2026-10-20 |
| LMArena Long Query | 1488 Elo | 1448 Elo | source · 2026-10-20 |
| LMArena Math | 1497 Elo | 1455 Elo | source · 2026-10-20 |
| LMArena Text | 1495 Elo | 1437 Elo | source · 2026-10-20 |
| LMArena Text Factuality | 1479 Elo | 1437 Elo | source · 2026-10-20 |
| LMArena Vision | 1290 Elo | 1242 Elo | source · 2026-10-19 |
| LMArena Vision Diagrams | 1322 Elo | 1270 Elo | source · 2026-10-19 |
| LMArena Vision OCR | 1307 Elo | 1262 Elo | source · 2026-10-19 |
| LMArena WebDev | 1659 Elo | 1593 Elo | source · 2026-10-19 |
| LMArena WebDev Image-to-Web | 1640 Elo | 1577 Elo | source · 2026-10-19 |
Values are as published by each evaluator, in its own units; rows from different evaluators are not on a common scale.
task evidence
How they place when the question changes
Lower ranks are better. “Best” marks the strongest option in this shortlist for that task.