Claude Opus 5 vs Muse Spark
Claude Opus 5 
Muse Spark Verdict
- Claude Opus 5 leads on quality: 70.0 vs 65.4.
- Claude Opus 5 is stronger in agents & tools, coding, composite, human preference, knowledge, long context, maths, multimodal, reasoning.
- Muse Spark is stronger in instruction following.
| Metric | Claude Opus 5 | Muse Spark |
|---|---|---|
| BenchLeader Index | 70.0 | 65.4 |
| Agents & tools score | 71.6 | 66.3 |
| Coding score | 74.4 | 66.0 |
| Composite score | 92.9 | 68.4 |
| Human preference score | 78.2 | 69.0 |
| Knowledge score | 70.5 | 64.5 |
| Long context score | 66.2 | 65.5 |
| Maths score | 67.1 | 56.3 |
| Multimodal score | 69.4 | 65.9 |
| Reasoning score | 71.1 | 70.0 |
| Instruction following score | – | 79.6 |
| Blended price $/M | $10.00 | – |
| Output speed | 52 tok/s | – |
| Time to first answer | 81.6 s | – |
| Context window | 1M | 262k |
| GPQA Diamond | 92.9% | 89.8% |
| OTIS Mock AIME | 97.8% | 88.9% |
| Humanity's Last Exam | – | 40.6% |
| SimpleBench | 80.6% | – |
| SciCode | – | 51.5% |
| WeirdML | 86.3% | – |
| ProofBench | – | 17.0% |
| Epoch Capabilities Index | 162.6 | 152.1 |
| LMArena Text | – | 1488 |
| LMArena Hard Prompts | – | 1505 |
| LMArena Coding | – | 1526 |
| LMArena Vision | – | 1306 |
| AA Intelligence Index | 50.7 | 31.3 |
| IFBench | – | 75.9% |
| AA-LCR | 79.3% | 78.0% |
| MMMU-Pro | 84.7% | 80.5% |
| AA-Omniscience | 37.1 | 7.2 |
| Terminal-Bench Hard | – | 45.5% |
| GPQA Diamond (AA) | 93.2% | 88.4% |
| Humanity's Last Exam (AA) | 54.9% | 40.7% |
| SciCode (AA) | 56.4% | – |
| τ²-Bench Telecom (AA) | – | 91.5% |
| AIME (Vals) | – | 96.9% |
| LiveCodeBench | 89.0% | – |
| MMLU-Pro | 91.6% | 87.3% |
| IOI | 91.7% | – |
| LegalBench | 87.0% | 84.2% |
| CorpFin | 73.2% | 65.1% |
| TaxEval | 75.1% | 77.7% |
| Terminal-Bench 2.1 (Vals) | 84.6% | – |
| SWE-bench (Vals) | 97.0% | 74.4% |
| GPQA Diamond (Vals) | 93.4% | 89.7% |
| Vals Index | 67.2 | – |
| SWE-Bench Pro | – | 55.0% |
| MCP Atlas | – | 82.2% |
| MultiChallenge | – | 75.5% |
| PRBench Finance | – | 52.4% |
| PRBench Legal | – | 52.3% |
| MultiNRC | – | 59.0% |
| HiL-Bench | 57.0% | – |
| TutorBench | – | 68.5% |
| EQ-Bench 4 | 1385 | – |
Data as of 2026-09-09. Best configuration of each model; every score links to its source on the model pages.