Claude Opus 5 vs Kimi K3
Claude Opus 5 (Anthropic) and Kimi K3 (Moonshot AI) compared on price, context window, speed and independent benchmarks. Claude Opus 5 scores higher on the Artificial Analysis Intelligence Index (63.1 vs 59.7), while Kimi K3 is 1.7× cheaper on a blended basis. Pick Claude Opus 5 for the hardest tasks and Kimi K3 when cost dominates.
| Claude Opus 5 | Kimi K3 | |
|---|---|---|
| Provider | Anthropic | Moonshot AI |
| Input $/1M | $5.00 | $3.00 |
| Output $/1M | $25.0 | $15.0 |
| Blended $/1M (3:1) | $10.0 | $6.00 |
| Cached input $/1M | $0.50 | $0.30 |
| Context window | 1M | 1.0M |
| Knowledge cutoff | 2026-05 | — |
| License | proprietary | open weights |
| Speed (tok/s) | 52 | 38 |
| Artificial Analysis Intelligence Index | 63.1 | 59.7 |
| LMArena Elo | 1505 | 1476 |
| GPQA Diamond | 93.2% | 93.5% |
| SWE-bench Verified | 97% | 93.4% |
| LiveCodeBench | 74.6% | — |
| MMLU-Pro | 91.6% | — |
| Humanity's Last Exam | 54.9% | 46.9% |
| MMMU | — | 81.6% |