Kimi K2 Thinking

google-vertex/moonshotai/kimi-k2-thinking-maas · family kimi-thinking · price provider-declared (via models.dev), USD per 1M tokens. Capabilities: toolsreasoningtext.

$1.07Blended $/1M (3:1)

$0.6Input $/1M

$2.5Output $/1M

$3.101M in + 1M out

—Cache read $/1M

#3262 of 5,000Cheapest rank

Provider: Vertex (google-vertex)
Context: 262,144 tokens
Max output: 262,144 tokens
Capabilities: toolsreasoningtext

Same model, other providers

“Kimi K2 Thinking” is offered by 9 providers — cheapest-effective first. Vercel AI Gateway is the cheapest at $0.85 blended.

Provider	Input	Output	Blended	Context
Vercel AI Gateway	$0.47	$2	$0.85	216k
Helicone	$0.48	$2	$0.86	256k
OpenCode Zen	$0.4	$2.5	$0.93	262k
IO.NET	$0.55	$2.25	$0.98	33k
Vertex (this)	$0.6	$2.5	$1.07	262k
NanoGPT	$0.6	$2.5	$1.07	262k
LLM Gateway	$0.6	$2.5	$1.07	262k
ZenMux	$0.6	$2.5	$1.07	262k
Moonshot AI (China)	$0.6	$2.5	$1.07	262k

Call it

jishie indexes & prices models; it does not proxy inference. Most providers are OpenAI-compatible — point base_url at Vertex and pass this model id:

curl $BASE_URL/v1/chat/completions \
  -H "Authorization: Bearer $API_KEY" -H "Content-Type: application/json" \
  -d '{"model":"moonshotai/kimi-k2-thinking-maas","messages":[{"role":"user","content":"hello"}]}'

Route programmatically with MCP find_model / get_model, or fetch this record free (no wallet):