Kimi K3 是 Kimi 迄今能力最强的旗舰模型,拥有 2.8 万亿参数,基于 KDA 混合线性注意力机制(Kimi Delta Attention)和注意力残差(Attention Residuals)技术构建,原生支持视觉理解,并拥有 100 万 token 上下文窗口。它是全球首个开源的 3 万亿级别模型,面向长程编程、知识工作和推理等前沿智能场景而设计。
from openai import OpenAI
client = OpenAI(
base_url="https://api.openroute.cn/v1",
api_key="<ONTOROUTE_API_KEY>",
)
completion = client.chat.completions.create(
model="kimi-k3",
messages=[
{
"role": "user",
"content": "Write a one-sentence bedtime story about a unicorn.",
}
],
)
print(completion.choices[0].message.content)Prices are CNY (¥) per million tokens, billed as you go. Cache reads can cut the cost of repeated context.
| Provider | Max Input | Max Output | Context length | Tool Calling |
|---|---|---|---|---|
moonshot moonshot/kimi-k3 | 1.048576M | 256K | 1.048576M | Supported |
This model does not publish tunable parameters.