Kimi K3 vs Muse Spark 1.2
Kimi K3 
Muse Spark 1.2 Verdict
- Kimi K3 leads on quality: 66.5 vs 62.6.
- Kimi K3 is stronger in agents & tools, human preference, long context, maths, multimodal.
- Muse Spark 1.2 is stronger in coding, composite, knowledge, reasoning.
- Muse Spark 1.2 is 3.0× cheaper ($2.00 vs $6.00 per 1M blended).
- Muse Spark 1.2 streams 6.0× faster (219 vs 37 tokens per second).
| Metric | Kimi K3 | Muse Spark 1.2 |
|---|---|---|
| BenchLeader Index | 66.5 | 62.6 |
| Agents & tools score | 67.5 | – |
| Coding score | 64.9 | 66.5 |
| Composite score | 74.1 | 79.1 |
| Human preference score | 71.2 | – |
| Knowledge score | 68.1 | 76.9 |
| Long context score | 71.1 | 66.0 |
| Maths score | 78.1 | 54.3 |
| Multimodal score | 65.1 | – |
| Reasoning score | – | 70.2 |
| Blended price $/M | $6.00 | $2.00 |
| Output speed | 37 tok/s | 219 tok/s |
| Time to first answer | 57.7 s | 23.7 s |
| Context window | 1.0M | 1.0M |
| SimpleBench | – | 74.5% |
| SciCode | 58.7% | – |
| APEX-Agents | 39.3% | – |
| FrontierCode | 44.2% | – |
| ProofBench | 87.0% | 43.0% |
| Epoch Capabilities Index | 157.6 | 155.5 |
| LiveBench | 79.2% | – |
| LiveBench Reasoning | 90.7% | – |
| LiveBench Coding | 81.5% | – |
| LiveBench Agentic Coding | 62.2% | – |
| LiveBench Mathematics | 84.4% | – |
| LiveBench Data Analysis | 78.7% | – |
| LiveBench Language | 85.5% | – |
| AA Intelligence Index | 43.8 | 39.8 |
| AA-LCR | 88.7% | 79.0% |
| MMMU-Pro | 80.5% | – |
| AA-Omniscience | 19.7 | 27.2 |
| GPQA Diamond (AA) | 93.5% | 90.4% |
| Humanity's Last Exam (AA) | 46.9% | 45.5% |
| SciCode (AA) | 59.5% | 57.4% |
| IOI | – | 49.5% |
| LegalBench | 86.0% | – |
| CorpFin | 71.6% | – |
| TaxEval | 75.7% | – |
| Terminal-Bench 2.1 (Vals) | 80.9% | – |
| SWE-bench (Vals) | 93.4% | – |
| GPQA Diamond (Vals) | 92.9% | – |
| Vals Index | 57.8 | – |
| EQ-Bench 4 | 1339 | – |
Data as of 2026-09-09. Best configuration of each model; every score links to its source on the model pages.