Up to four models
Kimi K3 vs Muse Spark 1.1
On the index (GPT-4 = 100), Kimi K3 scores 168 and Muse Spark 1.1 161. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.
Kimi K3
168152–183
WeightsOpen
ReleasedJul 2026
Price in / out$0.19 / $16
Context1.05M
Core tests taken6 of 9
161145–177
WeightsClosed
ReleasedJul 2026
Price in / out$1.25 / $4.25
Context1.05M
Core tests taken2 of 9
Kimi K3 leads Muse Spark 1.1 by 7 points, ahead on 1 of 2 shared core tests, at 2.1× the price.
Test by testthird-party results
Kimi K3
Muse Spark 1.1
Every result as a table
| Test | Kimi K3 | Muse Spark 1.1 |
|---|---|---|
| Reasoning & knowledge | ||
| LiveBench Data Analysis | 78.7% | 72.5% |
| LiveBench IF | 71.4% | 69.6% |
| LiveBench Language | 85.5% | 74.3% |
| LiveBench Reasoning | 90.7% | 87.7% |
| GPQA Diamond (Vals AI) | 92.9% | 91.2% |
| MMLU Pro | 88.0% | 88.7% |
| MMMU Pro | 88.1% | 86.6% |
| SimpleQA Verified | 50.6% | 57.8% |
| Maths | ||
| LiveBench Mathematics | 84.4% | 87.1% |
| Chess puzzles | 37.0% | 34.0% |
| Surface Evolver Bench | 95.0% | 52.5% |
| Coding | ||
| LiveBench Agentic Coding | 62.2% | 58.5% |
| LiveBench Coding | 81.5% | 77.2% |
| Code Migration | 16.1% | 31.1% |
| LiveCodeBench | 87.2% | 85.9% |
| ProgramBench | 2.0% | 0.0% |
| SWE-bench | 93.4% | 82.0% |
| Vibe Code Bench v1.1 | 85.0% | 72.2% |
| DeepSWE | 68.5% | 53.3% |
| Agents | ||
| APEX-Agents | 47.8% | 31.8% |
| DTBench | 91.2% | 94.4% |
| LMCA | 52.7% | 49.9% |
| Professional work | ||
| Finance Agent (v2) | 53.1% | 57.2% |
| CorpFin v2 | 71.6% | 71.3% |
| EMB | 66.7% | 56.4% |
| LegalBench | 86.2% | 85.0% |
| Legal Research Bench | 46.2% | 38.0% |
| MedScribe | 88.0% | 88.9% |
| MortgageTax | 66.3% | 66.1% |
| Public Benefits Bench v1.1 | 68.2% | 61.3% |
| TaxEval v2 | 75.7% | 79.7% |
Tests at least two of these models have taken. Bold rows are the core tests.