KIMI / API PRICING
kimi-k2-0905-preview API pricing
This page documents the gateway call identifier kimi-k2-0905-preview. It is not a claim that APIKIT is the model provider or that this routing name is a newly released official version.
Data modified: 2026-10-24T05:46:29Z
Public group prices
| Group / tier | Input / 1M | Output / 1M | Cache read / 1M | Rate |
|---|---|---|---|---|
| Kimi月之暗面Base | $0.196 | $0.784 | $0.049 | 0.35× |
Prices come from the public gateway configuration, not an independently verified provider retail quote. The configured convention is RMB 1 of balance counts as USD 1 in platform accounting; this is not a currency-exchange rate. Actual access, billing and discounts depend on your account.
A 1M-token cost example
Using the base tier of Kimi月之暗面, across multiple requests that each stay within the base pricing tier, cumulative 1 million uncached input tokens + 1 million output tokens = $0.196 + $0.784 = $0.98. This is not a quote for a single long-context request. This excludes cache, images, video, long-context tiers and time-dependent multipliers.
Integration examples
Use a key authorized for this group. Set the environment variable on your server; examples do not grant access or send a paid request by themselves.
Python
import json, os, urllib.request
url = 'https://www.apikit.cc/v1/chat/completions'
headers = {'Authorization': 'Bearer YOUR_API_KEY', 'Content-Type': 'application/json'}
# Keep APIKIT_API_KEY on the server
headers = {k: v.replace("YOUR_API_KEY", os.environ["APIKIT_API_KEY"]) for k, v in headers.items()}
payload = {'model': 'kimi-k2-0905-preview', 'messages': [{'role': 'user', 'content': 'Hello, APIKIT.'}], 'max_tokens': 256}
request = urllib.request.Request(url, data=json.dumps(payload).encode(), headers=headers, method="POST")
with urllib.request.urlopen(request) as response:
print(response.read().decode())Node.js
const key = process.env.APIKIT_API_KEY;
if (!key) throw new Error("Set APIKIT_API_KEY first");
const headers = {
"Authorization": "Bearer YOUR_API_KEY",
"Content-Type": "application/json"
};
for (const name in headers) headers[name] = headers[name].replace("YOUR_API_KEY", key);
const response = await fetch("https://www.apikit.cc/v1/chat/completions", {
method: "POST",
headers,
body: JSON.stringify({
"model": "kimi-k2-0905-preview",
"messages": [
{
"role": "user",
"content": "Hello, APIKIT."
}
],
"max_tokens": 256
})
});
if (!response.ok) throw new Error(`HTTP ${response.status}: ${await response.text()}`);
console.log(await response.json());cURL
curl 'https://www.apikit.cc/v1/chat/completions' \
-H 'Authorization: Bearer YOUR_API_KEY' \
-H 'Content-Type: application/json' \
--data-raw '{"model": "kimi-k2-0905-preview", "messages": [{"role": "user", "content": "Hello, APIKIT."}], "max_tokens": 256}'Parameters and billing boundaries
Context window, maximum output and capability limits are not supplied in the public catalog. Pricing intervals such as 200K or 272K are billing thresholds, not a guaranteed context limit.
Cache and scheduled rates may change the final charge. Missing prices are shown as an em dash, never as free usage.