kimi-k3 vs kimi-k3.1-pro-highspeed
| kimi-k3 | kimi-k3.1-pro-highspeed | |
|---|---|---|
| Providers | Moonshot AI | Moonshot AI |
| Input $/M | $1.54 | $2.00 |
| Output $/M | $7.69 | $10.46 |
| Output multiplier | ×5.00 | ×5.23 |
| Context window | 1M tokens | 1M tokens |
| Intelligence index | 43.6 | — |
| Coding index | 76.2 | — |
What the numbers mean
Input pricing is close — 1.30× apart — so the decision should come down to measured quality on your own evaluation set rather than headline cost.
Tags: kimi-k3 — Chat, Vision, Reasoning, Coding, 工具调用, Long context, 超大输出 · kimi-k3.1-pro-highspeed — Chat, Vision, Reasoning, Coding, 工具调用, Long context, 超大输出, 快速
Code
from openai import OpenAI
client = OpenAI(base_url="https://starseaapi.com/v1", api_key="sk-...")
# kimi-k3
client.chat.completions.create(model="kimi-k3", messages=[{"role":"user","content":"Hi"}])
# kimi-k3.1-pro-highspeed
client.chat.completions.create(model="kimi-k3.1-pro-highspeed", messages=[{"role": "user", "content": "Hi"}])