Muse Spark vs Muse Spark 1.1
Muse Spark Verdict
- Muse Spark and Muse Spark 1.1 are level on quality (65.4 vs 64.9).
- Muse Spark is stronger in agents & tools, human preference, instruction following, long context, maths, multimodal, reasoning.
- Muse Spark 1.1 is stronger in coding, composite, knowledge.
| Metric | Muse Spark | Muse Spark 1.1 |
|---|---|---|
| BenchLeader Index | 65.4 | 64.9 |
| Agents & tools score | 66.3 | 62.7 |
| Coding score | 66.0 | 68.2 |
| Composite score | 68.4 | 72.2 |
| Human preference score | 69.0 | 66.0 |
| Instruction following score | 79.6 | 77.8 |
| Knowledge score | 64.5 | 73.1 |
| Long context score | 65.5 | 65.3 |
| Maths score | 56.3 | 52.2 |
| Multimodal score | 65.9 | 64.9 |
| Reasoning score | 70.0 | 68.8 |
| Blended price $/M | – | $2.00 |
| Output speed | – | 194 tok/s |
| Time to first answer | – | 2.0 s |
| Context window | 262k | 1.0M |
| GPQA Diamond | 89.8% | – |
| OTIS Mock AIME | 88.9% | – |
| SimpleQA Verified | – | 57.8% |
| Humanity's Last Exam | 40.6% | – |
| SciCode | 51.5% | 58.2% |
| APEX-Agents | – | 41.9% |
| ProofBench | 17.0% | 39.0% |
| Epoch Capabilities Index | 152.1 | 154.6 |
| LMArena Text | 1488 | 1492 |
| LMArena Hard Prompts | 1505 | 1511 |
| LMArena Coding | 1526 | 1531 |
| LMArena WebDev | – | 1541 |
| LMArena Vision | 1306 | 1293 |
| LMArena Agent | – | -2.6 |
| AA Intelligence Index | 31.3 | 34.3 |
| IFBench | 75.9% | – |
| AA-LCR | 78.0% | 77.7% |
| MMMU-Pro | 80.5% | – |
| AA-Omniscience | 7.2 | 28.1 |
| Terminal-Bench Hard | 45.5% | – |
| GPQA Diamond (AA) | 88.4% | 89.8% |
| Humanity's Last Exam (AA) | 40.7% | 46.2% |
| SciCode (AA) | – | 58.8% |
| τ²-Bench Telecom (AA) | 91.5% | – |
| AIME (Vals) | 96.9% | – |
| MMLU-Pro | 87.3% | – |
| LegalBench | 84.2% | – |
| CorpFin | 65.1% | – |
| TaxEval | 77.7% | – |
| SWE-bench (Vals) | 74.4% | – |
| GPQA Diamond (Vals) | 89.7% | – |
| SWE-Bench Pro | 55.0% | 61.5% |
| MCP Atlas | 82.2% | 88.1% |
| MultiChallenge | 75.5% | 75.3% |
| PRBench Finance | 52.4% | 55.0% |
| PRBench Legal | 52.3% | 57.0% |
| MultiNRC | 59.0% | 65.6% |
| TutorBench | 68.5% | – |
| 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.