Getting Started

Follow this page to complete one minimal API call. The order is: sign in, top up, create an API key, copy a model name, and make a test request.

Before You Begin

Confirm the following before starting:

Item	Check
Tokeness account	Registered and signed in to tokeness.io
Balance	Top up a small test balance first
Model name	Copy one available model name from the model marketplace
Test method	Pick one: client, Node.js, Python, or cURL

If you plan to test directly with code:

Language	Install
Node.js	npm install openai
Python	pip install openai

1. Sign In to Tokeness

Open tokeness.io and click "Sign In" in the top right. When already signed in, you will go straight to the console.

Console after signing in

If this is your first time, register an account first. The email is used for sign-in, password recovery, and notifications.

2. Top Up Balance

After entering the console, open the "Wallet" page to top up. Tokeness uses a prepaid model; API calls are deducted from your balance.

Check before topping up:

Item	Check
Top-up account	Confirm you are signed in to the Tokeness account that will use the API
Top-up amount	Top up a small test amount first, then scale up based on your workload
Payment popup	If the browser blocks popups, allow Tokeness to open the payment page

3. Create an API Key

In the left navigation, go to "API Keys" and create a new key.

Split keys by project:

Scenario	Configuration
Development and testing	Separate key, small quota, easy to disable
Production service	Separate key, set clear quota and model scope
Customer project	One key per customer or project, for tracking consumption

Save the key immediately after creation. For security reasons, the documentation does not show screenshots of the real key page, and you should never share keys with untrusted parties.

4. Copy a Model Name and Configure the Base URL

Open the "Model Marketplace" and copy the model name you want to use. Copy the full model name exactly, preserving case, hyphens, and version numbers.

Endpoint address:

Common configuration:

Base URL: https://n.tokeness.io/v1
API Key: your Tokeness API key
Model ID: model name copied from the model marketplace

If a client splits the field into "Base URL / Endpoint / Service URL", put https://n.tokeness.io/v1 in all of them.

https://n.tokeness.io/v1

5. Make a Test Request

Below is a minimal Node.js example. Create test-tokeness.mjs first, then replace YOUR_TOKENESS_API_KEY and YOUR_MODEL_NAME with your own values.

import OpenAI from 'openai'

const client = new OpenAI({
  apiKey: 'YOUR_TOKENESS_API_KEY',
  baseURL: 'https://n.tokeness.io/v1'
})

const response = await client.chat.completions.create({
  model: 'YOUR_MODEL_NAME',
  messages: [{ role: 'user', content: 'Introduce Tokeness in one sentence.' }]
})

console.log(response.choices[0]?.message?.content)

Run:

npm install openai
node test-tokeness.mjs

If the call fails, check four things first: whether the key was copied completely, whether the account balance is sufficient, whether the model name matches the model marketplace, and whether the Base URL is https://n.tokeness.io/v1.