Kimi K3 API Resources

Kimi K3 API Resources

This page collects the practical Kimi K3 API guides, examples, and documentation links maintained around lizh.ai. Use it as a starting point when you want to test Kimi K3 through an OpenAI-compatible API endpoint.

lizh.ai is an independent OpenAI-compatible API gateway. It is not the official Moonshot AI or Kimi API website.

Core setup values

Start here

If you only need one link, open the public documentation mirror. It contains cURL, Python, Node.js, Cursor, Hermes, OpenClaw, pricing checks, and troubleshooting:

Kimi K3 OpenAI-compatible API Guide

Shop guides

These shop.lizh.ai guides cover high-intent search questions and tool-specific setup paths:

External examples and articles

What to test first

  1. Create an API key at lizh.ai/keys.
  2. Send one short cURL request to https://lizh.ai/v1/chat/completions.
  3. Move the same Base URL, API key, and model ID into Python, Cursor, Hermes, OpenClaw, or your backend.
  4. Check pricing and usage logs before long-context coding or agent workflows.

Common mistakes

  • Using a provider-specific Base URL instead of https://lizh.ai/v1.
  • Typing a display name instead of the exact model ID kimi-k3.
  • Testing a huge repository or long document before confirming a small request works.
  • Hard-coding API keys in source code.
  • Comparing model quality without tracking input tokens, output tokens, cache-hit tokens, latency, and cost.

Official reference

For official Kimi model behavior and Moonshot-specific features, read the provider documentation:

Kimi K3 Quickstart