GPT-5.6 Luna vs Kimi K3
GPT-5.6 Luna 
Kimi K3 Verdict
- Kimi K3 leads on quality: 66.5 vs 59.4.
- GPT-5.6 Luna (xhigh) is stronger in reasoning.
- Kimi K3 is stronger in agents & tools, coding, composite, human preference, knowledge, long context, multimodal, maths.
- GPT-5.6 Luna (xhigh) is 13× cheaper ($0.450 vs $6.00 per 1M blended).
- GPT-5.6 Luna (xhigh) streams 2.8× faster (107 vs 38 tokens per second).
| Metric | GPT-5.6 Luna (xhigh) | Kimi K3 |
|---|---|---|
| BenchLeader Index | 59.4 | 66.5 |
| Agents & tools score | 50.1 | 67.4 |
| Coding score | 60.0 | 65.0 |
| Composite score | 72.9 | 74.2 |
| Human preference score | 65.0 | 71.2 |
| Knowledge score | 59.0 | 68.2 |
| Long context score | 67.4 | 71.1 |
| Multimodal score | 61.6 | 65.0 |
| Reasoning score | 53.9 | – |
| Maths score | – | 78.1 |
| Blended price $/M | $0.450 | $6.00 |
| Output speed | 107 tok/s | 38 tok/s |
| Time to first answer | 55.4 s | 56.3 s |
| Context window | 1M | 1.0M |
| SimpleBench | 46.8% | – |
| SciCode | 50.0% | 58.7% |
| APEX-Agents | – | 39.3% |
| FrontierCode | – | 44.2% |
| ProofBench | – | 87.0% |
| Epoch Capabilities Index | – | 157.6 |
| LMArena Text | 1453 | – |
| LMArena Hard Prompts | 1473 | – |
| LMArena Coding | 1497 | – |
| LMArena WebDev | 1520 | – |
| LMArena Vision | 1254 | – |
| LMArena Agent | 1.1 | – |
| 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 | 34.8 | 43.8 |
| AA-LCR | 81.7% | 88.7% |
| MMMU-Pro | 78.5% | 80.5% |
| AA-Omniscience | -10.8 | 19.7 |
| GPQA Diamond (AA) | 89.5% | 93.5% |
| Humanity's Last Exam (AA) | 37.0% | 46.9% |
| SciCode (AA) | 50.5% | 59.5% |
| 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 |
| ARC-AGI-1 | 87.7% | – |
| ARC-AGI-2 | 47.6% | – |
| ARC-AGI-3 | 0.0% | – |
Data as of 2026-09-10. Best configuration of each model; every score links to its source on the model pages.
GPT-5.6 Luna vs Kimi K3: questions
- Is GPT-5.6 Luna better than Kimi K3?
- Kimi K3 leads on quality: 66.5 vs 59.4. The BenchLeader Index combines every independent quality benchmark; Kimi K3 is ahead overall as of 2026-09-10, but check the category scores for your use.
- Is GPT-5.6 Luna better than Kimi K3 for coding?
- Kimi K3 scores higher in coding (65 vs 60 on the category index, where 50 is average).
- Is GPT-5.6 Luna better than Kimi K3 for agentic tasks?
- Kimi K3 scores higher in agentic tasks (67 vs 50 on the category index, where 50 is average).
- Which is cheaper, GPT-5.6 Luna or Kimi K3?
- GPT-5.6 Luna is cheaper: $0.450 against $6.00 per million tokens, blended at three input tokens per output token.
- Which is faster, GPT-5.6 Luna or Kimi K3?
- GPT-5.6 Luna streams faster: 107 against 38 output tokens per second.
- Which has the larger context window?
- Kimi K3 accepts more context: 1.0M against 1M tokens.