Kimi K3.1 Pro API

Kimi K3.1 Pro is the successor to Kimi K3 from Moonshot AI. Reasoning is always on; you set its depth with reasoning_effort at three levels: low, high and max (default max). The context window is 1,048,576 tokens, shared by input and output. It supports tool calls (tools), JSON schema output and prompt caching. In Jantaris both input and output tokens cost less than on Kimi K3.

Two things to know up front. First, on follow-up turns send back the full assistant message including reasoning_content and tool_calls. Second, the default max can burn a lot of reasoning tokens - use low for simple tasks and cap length with max_tokens.

Best for

• Long coding sessions and tool-running agents
• Work over large repositories and documents (1M tokens)
• Tasks that need strict JSON output
• A cheaper model to benchmark against Claude for code

How to call it

from openai import OpenAI

client = OpenAI(base_url="https://jantaris.com/v1", api_key="sk-jt-...")
resp = client.chat.completions.create(
    model="kimi-k3.1-pro",
    reasoning_effort="low",
    max_tokens=4000,
    messages=[{"role": "user", "content": "Review this function for edge cases."}],
)
print(resp.choices[0].message.content)

FAQ

What is the Kimi K3.1 Pro API price?

$2.70 / $13.50 per 1M tokens. Cache reads cost less than regular input.

How is it different from Kimi K3?

It is the newer version at a lower token price. Test it on your own task: run it on Kimi K3 and K3.1 Pro, then compare output and cost in your dashboard.

How do I limit the answer length?

Use max_tokens and a lower reasoning_effort. Upstream, max_completion_tokens does not always stop the answer on this model.

How to call Kimi K3.1 Pro via API

from openai import OpenAI

client = OpenAI(base_url="https://jantaris.com/v1", api_key="sk-jt-...")
resp = client.chat.completions.create(
    model="kimi-k3.1-pro",
    messages=[{"role": "user", "content": "Write a haiku about Baltic amber."}],
)
print(resp.choices[0].message.content)

curl https://jantaris.com/v1/chat/completions \
  -H "Authorization: Bearer $JANTARIS_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model": "kimi-k3.1-pro", "messages": [{"role": "user", "content": "Write a haiku about Baltic amber."}]}'

import OpenAI from "openai";

const client = new OpenAI({ baseURL: "https://jantaris.com/v1", apiKey: process.env.JANTARIS_API_KEY });
const resp = await client.chat.completions.create({
  model: "kimi-k3.1-pro",
  messages: [{ role: "user", content: "Write a haiku about Baltic amber." }],
});
console.log(resp.choices[0].message.content);