IDUnify AI

One AI service.
Fair for every person.

A shared AI service for the websites in the IDUnify network. One account, one fair allowance per person, and no text stored.

  • By invitation
  • No questions or answers stored
  • Keys stay on your server
  • Works with OpenAI libraries
ai.idunify.comExample
POST /v1/chat
Authorization: Bearer idu_live_••••••••

{
  "feature": "demo.hello",
  "user": "u_7Fq2kM",
  "messages": [
    { "role": "user", "content": "Hello" }
  ]
}
200gemma4:31b0.8 credits
One call from your server
What you get

Everything a website needs to add AI.

Chat

Send a conversation. Get the answer at once, or streamed word by word.

Images

Add an image to a message as data or as a public link. We fetch links safely for you.

JSON answers

Ask for valid JSON only, ready for your code to read.

OpenAI-compatible

Already using an OpenAI library? Change only the base URL and the key.

Plans for your people

Create plans such as Free and Pro, move a person between them, or give bonus credits.

Usage you can see

Credits per project, feature, model and person, in your own dashboard.

How it works

From invitation to first answer.

  1. 1

    We invite you

    We set up your account with its limits and tell you what is included.

  2. 2

    You get your keys

    One key per project, sent to you separately. Keep each one on your server.

  3. 3

    Your server calls IDUnify AI

    Every call says who the person is, using your own id for them. No names or e-mail addresses.

  4. Credits are counted

    Usage is counted per person and per project. You see it in your dashboard.

For developers

One call. Your own user id.

The native call, and the same thing through an OpenAI library. Keys belong on your server only.

Native callcurl
curl https://ai.idunify.com/v1/chat \
  -H "Authorization: Bearer $IDUNIFY_KEY" \
  -H "Content-Type: application/json" \
  -d '{"feature":"demo.hello","user":"u_7Fq2kM",
       "messages":[{"role":"user","content":"Hello"}]}'
OpenAI libraryNode.js
import OpenAI from "openai";
const ai = new OpenAI({
  baseURL: "https://ai.idunify.com/v1",
  apiKey: process.env.IDUNIFY_KEY,
});
const r = await ai.chat.completions.create({
  messages: [{ role: "user", content: "Hello" }],
  user: "u_7Fq2kM",
});

Invited websites also get a Playground and ready-made code for curl, Node.js, Python and PHP.

Credits and limits

A fair allowance for every person.

01

Credits

One credit is about 1,000 tokens. Every answer shows what it cost.

02

Daily and monthly

Each person has a daily limit and a monthly allowance, so one person cannot use it all.

03

Plans

You decide what each plan allows, which models it may use, and who is on it.

04

Notices

You are told when usage reaches 80% of your limit and when it is reached, and you can ask for more.

Your dashboard

See it. Manage it. In English or Thai.

Your team signs in with its IDUnify account and sees usage, keys, plans and end users. What each person can do depends on their role in your team.

  • Create, rename and revoke keys
  • Manage plans and move people between them
  • Usage by day, project, feature and model; download as CSV
  • Notices, and requests for more credits
Privacy and security

Counts, never content.

We keep counts only

Who, when, which model, how many tokens and credits. IDUnify does not store the text of questions and answers, or your images.

Keys stay on your server

A request that carries a key from a web page is refused. A key is shown once and can be replaced at once.

Private to each website

A website sees only its own people and usage. Nothing is shared between websites.

Model provider

To answer, the text is processed by the AI model provider that runs the model, under its own terms.

Help

Common questions

For website owners

Want IDUnify AI on your website?

IDUnify AI is open by invitation. Tell us about your site and what you would build, and we'll get back to you.

support@idunify.com