Kimi K3 API
New512K context · Released Aug 26, 2026 · By Moonshot AI · Modified MIT
Kimi K3 costs $1.85 per million input tokens and $9 per million output tokens, against an official rate of $3 and $15 — 40% below official pricing. It is a Moonshot AI model under the Modified MIT licence, with a 512K context window, callable through an OpenAI-compatible endpoint with the model id `moonshotai/kimi-k3`.
Input
$1.85/M
$3.00/M
Output
$9.00/M
$15.00/M
Savings
40%
off official
- No markup
- OpenAI SDK compatible
- No minimum spend
- Modified MIT
The current open-weight frontier. Half-million-token context and frontier-class reasoning, at roughly a fifth of what closed models charge for the same work.
Copy this and go
The model id is already filled in. Nothing in this snippet needs editing except your key.
from openai import OpenAI
client = OpenAI(
api_key="YOUR_API_KEY",
base_url="https://router.xark.io/api/v1" # ← the only line that changes
)
response = client.chat.completions.create(
model="moonshotai/kimi-k3",
messages=[{"role": "user", "content": "Hello"}]
)
Kimi K3 pricing across providers
Rates as published by each provider. We update these when they change; if you spot one that's stale, tell us and we'll fix it.
| Provider | Input | Output |
|---|---|---|
| AI Token Router | $1.85/M | $9.00/M |
| Official (Moonshot AI) | $3.00/M | $15.00/M |
| OpenRouter | $2.94/M | $14.70/M |
| DeepInfra | $2.40/M | $11.50/M |
| Together AI | Not offered | — |
Cached input on Kimi K3 is billed at $0.37/M — roughly 80% below the standard input rate. For agent loops with a large fixed system prompt this is usually the number that matters most.
What it's good at
- Whole-codebase reasoning without a retrieval layer
- Research synthesis over hundreds of documents
- Complex multi-step planning agents