Kimi K3 API pricing and code examples

Kimi K3 is a Moonshot model you can call through LivRouter's OpenAI-compatible endpoint. There is no separate Moonshot account and no separate invoice: one LivRouter API key reaches Kimi K3 alongside every other model on the platform, billed per token from prepaid credits.

Kimi K3 API at a glance

Compare Kimi K3 API pricing, model ID, and OpenAI-compatible examples through LivRouter.

Model ID
kimi-k3
Provider
Moonshot
API format
OpenAI-compatible
Billing
Per token from LivRouter credits
Common integration focus
Moonshot API integrations

Kimi K3 pricing

Rates are in USD per 1M tokens. The reference row is the provider base rate recorded in LivRouter's catalog; the plan rows show the published rates for each eligible plan.

PlanInput / 1MOutput / 1MCached input / 1M
Reference (Moonshot)$3.00$15.00$0.30
Start$3.00$15.00$0.30
Max Kimi−79%$0.63$3.15$0.063
Pro Kimi−76%$0.72$3.60$0.072

Lowest listed LivRouter rate for Kimi K3: $0.63 per 1M input tokens on the Max Kimi plan — about 79% below the Moonshot reference rate. See every plan and rate.

Source: LivRouter model catalog. Catalog checked .

What does a Kimi K3 request cost?

One million uncached input tokens plus 100,000 output tokens at the listed Max Kimi rates cost $0.945. This example excludes subscription fees and assumes the selected plan is active. Actual charges depend on input, output, and cached-token usage.

Compare plan fees and credit allowances

Call Kimi K3 with the OpenAI SDK

Kimi K3 follows the OpenAI chat completions format, so existing OpenAI client code keeps working — change the base URL and the model name.

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.livrouter.com/v1",
  apiKey: process.env.LIVROUTER_API_KEY,
});

const completion = await client.chat.completions.create({
  model: "kimi-k3",
  messages: [{ role: "user", content: "Hello" }],
});

console.log(completion.choices[0].message.content);

Prefer plain HTTP? The same request over curl:

curl https://api.livrouter.com/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer $LIVROUTER_API_KEY" \
  -d '{
    "model": "kimi-k3",
    "messages": [{"role": "user", "content": "Hello"}]
  }'

Frequently asked questions about Kimi K3

How much does kimi-k3 cost through LivRouter?

kimi-k3 is metered per token. The Moonshot reference rate is $3.00 per 1M input tokens and $15.00 per 1M output tokens. On the Max Kimi plan LivRouter charges $0.63 per 1M input tokens, about 79% below the Moonshot reference rate. Cached input is $0.063 per 1M tokens.

Can I call kimi-k3 with the OpenAI SDK?

Yes. kimi-k3 is reachable through the OpenAI-compatible endpoint at https://api.livrouter.com/v1. Keep your existing OpenAI client code and change the base URL and the model name to "kimi-k3".

Do I need a separate Moonshot account?

No. LivRouter routes the request for you. You only need a LivRouter API key and credits — no separate Moonshot subscription.

Is kimi-k3 available on the Start tier?

Yes. kimi-k3 is reachable from the Start tier group, billed pay-as-you-go at the rates shown above.

Other Moonshot models on LivRouter

View all model pricingMoonshot API overviewQuickstart guide