Mi:dm K 3.0 Pro

by Kt

New

kt/midm-k-3.0-pro

Input: $0.28/1M tokens
Output: $1.12/1M tokens
Context: 262Ktokens
Max output: 236Ktokens
Knows to: -
Released: 5 Nov 2026

About

Mi:dm K 3.0 Pro is a Mixture-of-Experts reasoning model from KT, with 120B total and 12B active parameters. It uses a hybrid architecture that mixes Mamba, MoE and attention layers.

Pricing

Input	$0.28	per 1M prompt tokens
Output	$1.12	per 1M generated tokens, reasoning included
Cached input	$0.06	per 1M tokens read back from the prompt cache

What a request costs

Typical request	Each	Per 1,000
Short chat reply1,500 in · 400 out	$0.0009	$0.868
Question over a long document40,000 in · 800 out	$0.012	$12.10
Agent tool step20,000 in · 1,200 out	$0.0069	$6.94

At the developer-API rate, before any provider promotion or prompt-cache saving. In the app, the model picker shows what a message costs on your plan before you send it.

Call it from code

OpenAI-compatible: point any OpenAI SDK at https://api.privateer.pro/v1 and set model. Keys are created in the app under Settings.

cURL

curl https://api.privateer.pro/v1/chat/completions \
  -H "Authorization: Bearer $PRIVATEER_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "kt/midm-k-3.0-pro",
    "messages": [{ "role": "user", "content": "Hello" }]
  }'

Python

import os
from openai import OpenAI

client = OpenAI(
    base_url="https://api.privateer.pro/v1",
    api_key=os.environ["PRIVATEER_API_KEY"],
)
reply = client.chat.completions.create(
    model="kt/midm-k-3.0-pro",
    messages=[{"role": "user", "content": "Hello"}],
)
print(reply.choices[0].message.content)

JavaScript

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.privateer.pro/v1",
  apiKey: process.env.PRIVATEER_API_KEY,
});
const reply = await client.chat.completions.create({
  model: "kt/midm-k-3.0-pro",
  messages: [{ role: "user", content: "Hello" }],
});
console.log(reply.choices[0].message.content);

Parameters this model accepts

frequency_penaltyinclude_reasoninglogprobsmax_tokenspresence_penaltyreasoningrepetition_penaltyresponse_formatstructured_outputstemperaturetop_ktop_logprobstop_p

API reference → · How API billing works →

Where your prompt goes

No zero-data-retention route

None of the endpoints serving this model offer zero data retention. Privateer pins requests to zero-data-retention endpoints by default, so with the default setting a request to this model is refused rather than sent somewhere that may keep it.

Over the API, prompts are forwarded to the model provider for inference and are not encrypted in transit to it; we persist only billing metadata (timestamp, model ID, token counts, cost), never your prompts or the responses. In the app, your chats are encrypted on your device or in our cloud.

How the privacy tiers work →

Capabilities

Reads: Text
Writes: Text
Tool calling: No
Reasoning: Yes
Structured output: Yes
Prompt caching: Yes
Provider moderation: None

Reliability

Serving endpoints: 1
Uptime, last day: 100%best endpoint

Measured across the hosts serving this model, as of 11 Nov 2026. The router sends traffic to the healthy ones, so a single bad host rarely shows up in a request.

Prices are USD at the developer-API rate, as of 11 Nov 2026, rebuilt daily from the providers’ live catalogues. ← All models