GPT-5.6 Sol vs Muse Spark 1.1
GPT-5.6 Sol 
Muse Spark 1.1 Verdict
- GPT-5.6 Sol (high) leads on quality: 68.0 vs 64.9.
- GPT-5.6 Sol (high) is stronger in agents & tools, coding, composite, knowledge, long context, multimodal.
- Muse Spark 1.1 is stronger in instruction following, reasoning, human preference, maths.
- Muse Spark 1.1 is 4.0× cheaper ($2.00 vs $8.00 per 1M blended).
- Muse Spark 1.1 streams 3.3× faster (194 vs 59 tokens per second).
| Metric | GPT-5.6 Sol (high) | Muse Spark 1.1 |
|---|---|---|
| BenchLeader Index | 68.0 | 64.9 |
| Agents & tools score | 87.6 | 62.7 |
| Coding score | 72.0 | 68.2 |
| Composite score | 82.5 | 72.2 |
| Instruction following score | 67.6 | 77.8 |
| Knowledge score | 73.7 | 73.1 |
| Long context score | 67.4 | 65.3 |
| Multimodal score | 66.5 | 64.9 |
| Reasoning score | 62.9 | 68.8 |
| Human preference score | – | 66.0 |
| Maths score | – | 52.2 |
| Blended price $/M | $8.00 | $2.00 |
| Output speed | 59 tok/s | 194 tok/s |
| Time to first answer | 31.4 s | 2.0 s |
| Context window | 1M | 1.0M |
| SimpleQA Verified | – | 57.8% |
| SciCode | 56.9% | 58.2% |
| WeirdML | 88.8% | – |
| APEX-Agents | – | 41.9% |
| ProofBench | – | 39.0% |
| Epoch Capabilities Index | – | 154.6 |
| LMArena Text | – | 1492 |
| LMArena Hard Prompts | – | 1511 |
| LMArena Coding | – | 1531 |
| LMArena WebDev | – | 1541 |
| LMArena Vision | – | 1293 |
| LMArena Agent | – | -2.6 |
| AA Intelligence Index | 42.5 | 34.3 |
| IFBench | 69.2% | – |
| AA-LCR | 81.7% | 77.7% |
| MMMU-Pro | 81.8% | – |
| AA-Omniscience | 20.4 | 28.1 |
| Terminal-Bench Hard | 62.1% | – |
| GPQA Diamond (AA) | 92.8% | 89.8% |
| Humanity's Last Exam (AA) | 46.0% | 46.2% |
| SciCode (AA) | 57.8% | 58.8% |
| τ²-Bench Telecom (AA) | 83.3% | – |
| SWE-Bench Pro | – | 61.5% |
| MCP Atlas | – | 88.1% |
| MultiChallenge | – | 75.3% |
| PRBench Finance | – | 55.0% |
| PRBench Legal | – | 57.0% |
| EnigmaEval | 37.1% | – |
| MultiNRC | – | 65.6% |
| EQ-Bench 4 | – | 1260 |
| ARC-AGI-1 | 97.0% | – |
| ARC-AGI-2 | 85.4% | – |
| ARC-AGI-3 | 2.2% | – |
Data as of 2026-09-09. Best configuration of each model; every score links to its source on the model pages.