GPT-6 Astra vs Kimi K3
GPT-6 Astra 
Kimi K3 Verdict
- GPT-6 Astra (max) leads on quality: 70.6 vs 66.5.
- GPT-6 Astra (max) is stronger in agents & tools, coding, knowledge, reasoning.
- Kimi K3 is stronger in composite, maths, human preference, long context, multimodal.
- Kimi K3 is 3.7× cheaper ($6.00 vs $22.00 per 1M blended).
- GPT-6 Astra (max) streams 1.5× faster (54 vs 37 tokens per second).
| Metric | GPT-6 Astra (max) | Kimi K3 |
|---|---|---|
| BenchLeader Index | 70.6 | 66.5 |
| Agents & tools score | 68.2 | 67.5 |
| Coding score | 79.0 | 64.9 |
| Composite score | 73.1 | 74.1 |
| Knowledge score | 79.8 | 68.1 |
| Maths score | 77.0 | 78.1 |
| Reasoning score | 70.9 | – |
| Human preference score | – | 71.2 |
| Long context score | – | 71.1 |
| Multimodal score | – | 65.1 |
| Blended price $/M | $22.00 | $6.00 |
| Output speed | 54 tok/s | 37 tok/s |
| Time to first answer | 328.6 s | 57.7 s |
| Context window | 1.1M | 1.0M |
| GPQA Diamond | 95.8% | – |
| FrontierMath Tiers 1–3 | 93.7% | – |
| FrontierMath Tier 4 | 97.6% | – |
| OTIS Mock AIME | 100.0% | – |
| SimpleQA Verified | 75.6% | – |
| Terminal-Bench | 58.2% | – |
| SciCode | 56.5% | 58.7% |
| APEX-Agents | – | 39.3% |
| FrontierCode | 53.3% | 44.2% |
| ProofBench | – | 87.0% |
| Epoch Capabilities Index | – | 157.6 |
| LMArena WebDev | 1796 | – |
| LMArena Agent | 12.5 | – |
| LiveBench | 82.2% | 79.2% |
| LiveBench Reasoning | 92.7% | 90.7% |
| LiveBench Coding | 80.4% | 81.5% |
| LiveBench Agentic Coding | 57.3% | 62.2% |
| LiveBench Mathematics | 96.8% | 84.4% |
| LiveBench Data Analysis | 83.0% | 78.7% |
| LiveBench Language | 89.4% | 85.5% |
| AA Intelligence Index | – | 43.8 |
| AA-LCR | – | 88.7% |
| MMMU-Pro | – | 80.5% |
| AA-Omniscience | – | 19.7 |
| GPQA Diamond (AA) | – | 93.5% |
| Humanity's Last Exam (AA) | – | 46.9% |
| SciCode (AA) | – | 59.5% |
| LegalBench | – | 86.0% |
| CorpFin | – | 71.6% |
| TaxEval | – | 75.7% |
| Terminal-Bench 2.1 (Vals) | 87.3% | 80.9% |
| SWE-bench (Vals) | – | 93.4% |
| GPQA Diamond (Vals) | – | 92.9% |
| Vals Index | 66.6 | 57.8 |
| EQ-Bench 4 | – | 1339 |
| ARC-AGI-1 | 97.5% | – |
| ARC-AGI-2 | 95.0% | – |
| ARC-AGI-3 | 62.7% | – |
Data as of 2026-09-09. Best configuration of each model; every score links to its source on the model pages.