Kimi K2.6 is live on Lilac
By Lucas Ewing
TL;DR
Kimi K2.6 is now live on Lilac as an OpenAI-compatible shared endpoint.
| Model | Input | Cache read | Output | Context |
|---|---|---|---|---|
| Kimi K2.6 | $0.70 / M tokens | $0.20 / M tokens | $3.50 / M tokens | 262K |
No contracts, no reserved-capacity markup, no minimums. Use the same OpenAI SDK and point it at Lilac.
Why Kimi K2.6
Kimi K2.6 is Moonshot AI's latest Kimi model for coding, long-context execution, tool use, and agent workflows. The official Kimi launch page describes it as a model for building real products, coordinating agent work, and handling complex tasks with stronger reliability.
Read the official model page here: Kimi K2.6 by Moonshot AI.
Pricing
Kimi K2.6 on Lilac uses three token rates:
- Input — $0.70 / M fresh input tokens
- Cache read — $0.20 / M repeated input tokens served from cache
- Output — $3.50 / M generated tokens
Cache-read pricing matters for coding agents, tool-heavy workflows, and long-context chat sessions where most of the prompt context repeats across turns.
OpenAI-compatible API
from openai import OpenAI
client = OpenAI(
base_url="https://api.getlilac.com/v1",
api_key="lilac_sk_...",
)
response = client.chat.completions.create(
model="moonshotai/kimi-k2.6",
messages=[
{"role": "user", "content": "Build a TypeScript helper for this API."},
],
)
Availability
Kimi K2.6 is available now on Lilac's shared warm endpoints. Start from the inference section, check the pricing docs, or sign up in the Lilac console.