Moonshot AI — Coding Model

Kimi K3 API — 1M tokens context.

Kimi K3 is Moonshot AI's long-context model, served through IN2PETA with a 1M-token context window. It is built for very large pasted context — whole repositories, long documents and archives in a single request — and is available on the Pro plan.

Model ID

kimi-k3

$3.00Input / 1M
$15.00Output / 1M

Kimi K3 specifications

Model IDkimi-k3
VendorMoonshot AI
Usage tierTier 3
Context window1M tokens
QuantizationFP8
Input$3.00 per 1M tokens
Output$15.00 per 1M tokens
Cache read$0.30 per 1M tokens
Approx. output speed~100 tokens/sec

Legacy model IDs that resolve to Kimi K3: kimi-2.7. Aliases only resolve when your plan includes the model.

Why run Kimi K3 on IN2PETA

  • 1M-token context window
  • Moonshot AI long-context specialist
  • Handles very large pasted context
  • OpenAI-compatible endpoint — switch providers by changing one base URL, not by rewriting your client.

Calling the Kimi K3 API

Kimi K3 is served from an OpenAI-compatible gateway. Change your base URL and API key, and existing OpenAI clients work unchanged.

curl https://code.in2peta.com/v1/coding/chat/completions \
  -H "Authorization: Bearer $IN2PETA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "kimi-k3",
    "messages": [{"role": "user", "content": "Explain this function"}]
  }'

Set ANTHROPIC_BASE_URL to the same host to point Claude Code at the same gateway. See Getting Started for key setup, and Coding Endpoints for the full API reference.

Plans that include Kimi K3

FreeFree

Not included

Starter$2/mo

Not included

Go$5/mo

Not included

Pro$15/mo

Includes this model

Frequently asked questions

How do I access the Kimi K3 API?

Point any OpenAI-compatible client at https://code.in2peta.com and send requests to /v1/coding/chat/completions with "model": "kimi-k3". The gateway is OpenAI-compatible, so existing SDKs work by changing the base URL and API key.

How much does the Kimi K3 API cost?

Kimi K3 is $3 per million input tokens and $15 per million output tokens, with cache reads at $0.3 per million. There is no per-request fee on top of token usage.

Which IN2PETA plans include Kimi K3?

Kimi K3 is available on Pro ($15/mo). Requests for a model outside your plan return HTTP 403 with the allowed model list in the error body.

What is the context window of Kimi K3?

Kimi K3 serves a 1M tokens context window and runs FP8 quantised on IN2PETA.

Start with Kimi K3 today

Kimi K3 is available from the Pro plan at $15/month.