Muse Spark 1.1 vs Qwen3 8
Muse Spark 1.1 
Qwen3 8 Verdict
- Muse Spark 1.1 leads on quality: 64.9 vs 63.9.
- Muse Spark 1.1 is stronger in agents & tools, composite, instruction following, knowledge, maths, reasoning.
- Qwen3 8 (max) is stronger in coding, human preference, long context, multimodal.
- Muse Spark 1.1 is 1.3× cheaper ($2.00 vs $2.67 per 1M blended).
- Muse Spark 1.1 streams 4.9× faster (194 vs 39 tokens per second).
| Metric | Muse Spark 1.1 | Qwen3 8 (max) |
|---|---|---|
| BenchLeader Index | 64.9 | 63.9 |
| Agents & tools score | 62.7 | 56.1 |
| Coding score | 68.2 | 71.7 |
| Composite score | 72.2 | 70.8 |
| Human preference score | 66.0 | 67.9 |
| Instruction following score | 77.8 | – |
| Knowledge score | 73.1 | 62.3 |
| Long context score | 65.3 | 65.7 |
| Maths score | 52.2 | – |
| Multimodal score | 64.9 | 67.2 |
| Reasoning score | 68.8 | 67.7 |
| Blended price $/M | $2.00 | $2.67 |
| Output speed | 194 tok/s | 39 tok/s |
| Time to first answer | 2.0 s | 53.2 s |
| Context window | 1.0M | 1M |
| SimpleQA Verified | 57.8% | – |
| SciCode | 58.2% | – |
| APEX-Agents | 41.9% | – |
| ProofBench | 39.0% | – |
| Epoch Capabilities Index | 154.6 | 156.6 |
| LMArena Text | 1492 | 1480 |
| LMArena Hard Prompts | 1511 | 1502 |
| LMArena Coding | 1531 | 1520 |
| LMArena WebDev | 1541 | 1685 |
| LMArena Vision | 1293 | 1313 |
| LMArena Agent | -2.6 | 3.9 |
| 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% |
| AA Intelligence Index | 34.3 | 40.3 |
| AA-LCR | 77.7% | 78.3% |
| MMMU-Pro | – | 82.3% |
| AA-Omniscience | 28.1 | 3.4 |
| GPQA Diamond (AA) | 89.8% | 92.7% |
| Humanity's Last Exam (AA) | 46.2% | 43.0% |
| SciCode (AA) | 58.8% | 53.2% |
| LiveCodeBench | – | 87.8% |
| MMLU-Pro | – | 88.6% |
| IOI | – | 73.0% |
| LegalBench | – | 83.6% |
| CorpFin | – | 65.8% |
| TaxEval | – | 75.5% |
| Terminal-Bench 2.1 (Vals) | – | 67.4% |
| SWE-bench (Vals) | – | 85.6% |
| GPQA Diamond (Vals) | – | 93.7% |
| Vals Index | – | 51.8 |
| SWE-Bench Pro | 61.5% | – |
| MCP Atlas | 88.1% | – |
| MultiChallenge | 75.3% | – |
| PRBench Finance | 55.0% | – |
| PRBench Legal | 57.0% | – |
| MultiNRC | 65.6% | – |
| EQ-Bench 4 | 1260 | – |
Data as of 2026-09-09. Best configuration of each model; every score links to its source on the model pages.