Kimi K2.7 Code

Kimi K2.7 Code is Moonshot AI's coding-focused agentic model built on Kimi K2.6. It improves real-world long-horizon coding task completion, instruction following, and token efficiency while reducing thinking-token usage by approximately 30% versus Kimi K2.6.

MCP-Mark

81.1%

i
MCP Atlas

76.0%

i
Kimi Code Bench v2

62.0%

i
Program Bench

53.6%

i
Kimi Claw 24/7 Bench

46.9%

i
MLS-Bench Lite

35.1%

i

Pricing, uptime, and speed via OpenRouter — updated Jun 13, 2026, 08:25 PM.

Provider	Status	Input	Output	Limits	Uptime	Speed	Notes
Ambient	available	$0.75/Mtok cache $0.16/Mtok	$3.50/Mtok	262K tokens context 262K tokens max output	98% 5m 100.0%	890 ms p50 TTFT 85 tok/s p50	int4
Parasail	available	$0.75/Mtok cache $0.16/Mtok	$3.50/Mtok	262K tokens context 262K tokens max output	99%	1,703 ms p50 TTFT 38 tok/s p50	int4
AtlasCloud	available	$0.95/Mtok cache $0.16/Mtok	$4.00/Mtok	262K tokens context 262K tokens max output	98% 5m 100.0%	939 ms p50 TTFT 51 tok/s p50	int4
Cloudflare	available	$0.95/Mtok cache $0.19/Mtok	$4.00/Mtok	262K tokens context 262K tokens max output	99% 5m 100.0%	5,264 ms p50 TTFT 27 tok/s p50
Io Net	available	$0.95/Mtok cache $0.47/Mtok	$4.00/Mtok	262K tokens context 262K tokens max output	100.0%	4,235 ms p50 TTFT 47 tok/s p50	int4
Moonshot AI	available	$0.95/Mtok cache $0.19/Mtok	$4.00/Mtok	262K tokens context 262K tokens max output	100.0% 5m 100.0%	2,179 ms p50 TTFT 28 tok/s p50	int4
Novita	available	$0.95/Mtok cache $0.19/Mtok	$4.00/Mtok	262K tokens context 262K tokens max output	100.0%	3,212 ms p50 TTFT 30 tok/s p50