Claude Opus 4.8 vs Kimi K2 Thinking
Anthropic vs Moonshot AI — benchmarks, pricing, and capabilities side by side.
- •Kimi K2 Thinking has the higher intelligence index (73.7 vs 73.1)
- •Kimi K2 Thinking is cheaper ($0.60 vs $6.30 per 1M input)
- •Kimi K2 Thinking is faster
| Claude Opus 4.8 | Kimi K2 Thinking | |
|---|---|---|
| Intelligence index | 73.1 | 73.7 |
| Developer | Anthropic | Moonshot AI |
| Type | LLM | LLM |
| Access | — | Open weights |
| Context window | — | 262,144 tokens |
| Input price | $6.30 / 1M | $0.60 / 1M |
| Output price | $25.00 / 1M | $2.50 / 1M |
| Speed | 66 tok/s | 100 tok/s |
| Released | May 28, 2026 | November 6, 2025 |
| Parameters | — | 1T (32B active) |
| Input modalities | — | Text |
| Output modalities | — | Text |
Shared benchmarks
Claude Opus 4.8
Kimi K2 Thinking
AA-LCR
67.7
66.3
GPQA Diamond
92
84.5
Humanity’s Last Exam
45.7
22.3
SciCode
53.5
42.4
Terminal-Bench
58.3
31.1
τ²-bench
94.4
93