Installation

Install the official Cencori SDK for your preferred language. The SDK is meant to run on the server because it uses your Cencori project secret key.

TypeScript / Node.js

npm install cencori
# or: yarn add cencori
# or: pnpm add cencori

Python

pip install cencori

Go

go get github.com/cencori/cencori-go

Requirements

• Node.js 18+ for the TypeScript SDK.
• TypeScript 5+ is optional, but recommended.
• A Cencori project key from Project > API Keys.
• Provider access for the model you want to call. Use managed access if enabled, or add your own provider key in Project > Providers.

Environment Variables

# .env.local for Next.js, .env for most Node apps
CENCORI_API_KEY=csk_...

Do not expose a csk_... key through NEXT_PUBLIC_* variables or browser-side code.

Configuration

import { Cencori } from 'cencori';
 
const cencori = new Cencori({
  apiKey: process.env.CENCORI_API_KEY,
  baseUrl: 'https://cencori.com', // optional; this is the default
  headers: {
    'X-Trace-ID': 'req_123',
  },
});

The SDK automatically reads CENCORI_API_KEY if no apiKey is provided:

const cencori = new Cencori();

OpenAI-Compatible Base URL

For tools that already speak the OpenAI API, install nothing from Cencori and set:

Base URL: https://api.cencori.com/v1
API key:  your Cencori project key

In most OpenAI-compatible clients, the base URL should stop at /v1; the client appends /chat/completions.

TypeScript Support

The SDK is written in TypeScript and includes full type definitions:

import { Cencori, ChatRequest, ChatResponse } from 'cencori';
 
const cencori = new Cencori();
 
const request: ChatRequest = {
  model: 'gpt-4o',
  messages: [{ role: 'user', content: 'Hello' }],
};
 
const response: ChatResponse = await cencori.ai.chat(request);

Next Steps

See Add Cencori to an Existing Product for dashboard-to-code setup, or Making Your First Request to send a request with the SDK.