Coding LLM Benchmark

← Back to dashboard

Kimi K2 Thinking

Moonshot AI · China

Pricing & capacity

Input price
$0.40 / 1M tokens
Output price
$1.75 / 1M tokens
Context window
256K

Capabilities

Vision
No
Reasoning
Yes
Tool calls
Yes
Cursor
No
OpenRouter
Yes

Coding benchmarks

LiveCodeBench
83.1%
Aider Polyglot
59.1%
SWE-bench Verified
71.3%
BFCL (Tool use)
59.42%
Code Arena Elo
1442

Notes

LiveCode 83.1%, SWE 71.3% (Vellum Nov 2025). Arena #9 (kimi-k2.5-thinking, Feb 2026). Very competitive pricing.