Grok 4.5 vs Kimi K2.6
Grok 4.5 
Kimi K2.6 Verdict
- Grok 4.5 leads on quality: 63.5 vs 60.7.
- Grok 4.5 is stronger in agents & tools, human preference, knowledge, multimodal, reasoning.
- Kimi K2.6 is stronger in coding, composite, long context, instruction following, maths.
- Kimi K2.6 is 1.8× cheaper ($1.71 vs $3.00 per 1M blended).
| Metric | Grok 4.5 | Kimi K2.6 |
|---|---|---|
| BenchLeader Index | 63.5 | 60.7 |
| Agents & tools score | 58.6 | 45.1 |
| Coding score | 59.8 | 60.4 |
| Composite score | 66.2 | 68.6 |
| Human preference score | 67.2 | 60.8 |
| Knowledge score | 76.2 | 58.6 |
| Long context score | 66.2 | 67.1 |
| Multimodal score | 64.9 | 63.7 |
| Reasoning score | 69.2 | 66.0 |
| Instruction following score | – | 73.7 |
| Maths score | – | 53.8 |
| Blended price $/M | $3.00 | $1.71 |
| Output speed | 52 tok/s | 44 tok/s |
| Time to first answer | 13.6 s | 103.2 s |
| Context window | 500k | 262k |
| GPQA Diamond | – | 90.8% |
| FrontierMath Tiers 1–3 | – | 57.2% |
| FrontierMath Tier 4 | – | 25.6% |
| OTIS Mock AIME | – | 96.1% |
| SWE-bench Verified (Epoch) | – | 76.7% |
| SimpleQA Verified | – | 34.9% |
| SimpleBench | 70.0% | – |
| OSWorld-Verified 2.0 | – | 4.6% |
| SciCode | – | 53.5% |
| WeirdML | 46.4% | 55.9% |
| APEX-Agents | 34.2% | 18.9% |
| FrontierCode | 42.4% | – |
| ProofBench | – | 16.0% |
| Epoch Capabilities Index | 153.9 | 151.0 |
| LMArena Text | 1471 | 1461 |
| LMArena Hard Prompts | 1495 | 1485 |
| LMArena Coding | 1523 | 1514 |
| LMArena WebDev | 1556 | 1509 |
| LMArena Vision | 1291 | 1281 |
| LMArena Agent | 3.9 | – |
| LiveBench | 75.8% | – |
| LiveBench Reasoning | 87.2% | – |
| LiveBench Coding | 68.6% | – |
| LiveBench Agentic Coding | 56.5% | – |
| LiveBench Mathematics | 90.8% | – |
| LiveBench Data Analysis | 73.0% | – |
| LiveBench Language | 82.8% | – |
| AA Intelligence Index | 39.1 | 31.3 |
| IFBench | – | 76.0% |
| AA-LCR | 79.3% | 81.0% |
| MMMU-Pro | 80.4% | 79.4% |
| AA-Omniscience | 25.3 | 5.3 |
| Terminal-Bench Hard | – | 43.9% |
| GPQA Diamond (AA) | 93.1% | 91.1% |
| Humanity's Last Exam (AA) | 42.7% | 37.5% |
| SciCode (AA) | 55.0% | 51.5% |
| τ²-Bench Telecom (AA) | – | 95.9% |
| LiveCodeBench | – | 86.8% |
| MMLU-Pro | – | 87.6% |
| LegalBench | – | 84.7% |
| CorpFin | – | 66.7% |
| TaxEval | – | 74.7% |
| Terminal-Bench 2.1 (Vals) | – | 53.6% |
| SWE-bench (Vals) | – | 76.2% |
| GPQA Diamond (Vals) | – | 89.1% |
| Vals Index | – | 43.5 |
| HiL-Bench | – | 18.7% |
| EQ-Bench 4 | – | 1202 |
| Kagi LLM Benchmark | 83.5% | – |
Data as of 2026-09-10. Best configuration of each model; every score links to its source on the model pages.
Grok 4.5 vs Kimi K2.6: questions
- Is Grok 4.5 better than Kimi K2.6?
- Grok 4.5 leads on quality: 63.5 vs 60.7. The BenchLeader Index combines every independent quality benchmark; Grok 4.5 is ahead overall as of 2026-09-10, but check the category scores for your use.
- Is Grok 4.5 better than Kimi K2.6 for coding?
- Kimi K2.6 scores higher in coding (60 vs 60 on the category index, where 50 is average).
- Is Grok 4.5 better than Kimi K2.6 for agentic tasks?
- Grok 4.5 scores higher in agentic tasks (59 vs 45 on the category index, where 50 is average).
- Which is cheaper, Grok 4.5 or Kimi K2.6?
- Kimi K2.6 is cheaper: $1.71 against $3.00 per million tokens, blended at three input tokens per output token.
- Which is faster, Grok 4.5 or Kimi K2.6?
- Grok 4.5 streams faster: 52 against 44 output tokens per second.
- Which has the larger context window?
- Grok 4.5 accepts more context: 500k against 262k tokens.