Nodexi Agentics
Home/ Documentation

Documentation

Everything you need to make your first request and integrate AI capabilities into your application. New to Nodexi? Start with the three steps below.

Getting started

Integrate AI in three steps

From first request to production, the path is simple and stays in your control.

1

Connect

Create an account and provision an API key for your environment. Keep your key secret and store it as an environment variable.

2

Integrate

Call the unified API directly over HTTP, or use the SDK for your language. Adapt the examples below to your use case.

3

Scale

Move to production and choose a plan that matches your traffic. Usage scales automatically as your product grows.

Authentication

Authenticate with an API key

Every request is authenticated with a bearer token — your API key. Provision separate keys per environment (development, staging, production) and rotate them whenever you need to.

  • Send your key in the Authorization header as a bearer token.
  • Never expose keys in client-side code or commit them to source control.
  • Rotate or revoke keys at any time from your account.
auth.sh
# Store your key as an environment variable
export NODEXI_KEY="sk_live_..."

# Send it as a bearer token on every request
curl https://api.nodexi.top/v1/complete \
  -H "Authorization: Bearer $NODEXI_KEY"
Your first request

Make a completion request

Send an input describing the task and optional context. The API returns a structured response with the generated output you can use directly in your product.

  • input — the instruction or task to perform.
  • context — optional supporting text the model should use.
  • output — the generated result in the response body.
example.js
import { Nodexi } from "@nodexi/sdk";

const client = new Nodexi({
  apiKey: process.env.NODEXI_KEY,
});

const response = await client.complete({
  input: "Summarize this support ticket",
  context: ticket.body,
});

console.log(response.output);
Response format

Predictable, structured responses

Responses share a consistent shape so you can parse them the same way across capabilities. Each response includes the generated output and metadata about the request.

response.json
{
  "id": "req_8f2c...",
  "output": "The customer reports...",
  "model": "nodexi-default",
  "usage": {
    "input_units": 128,
    "output_units": 42
  }
}
Reference

Keep going

Deeper reference material is being expanded. Reach out anytime and our team will point you to what you need.

API reference

Endpoints, parameters, and response fields for the unified API.

SDKs

Language SDKs that wrap the API with idiomatic helpers.

Guides

End-to-end walkthroughs for common integration patterns.

Support

Email our team and we'll help you get unblocked.

Talk to our team