Quickstart

This guide assumes your Deeplinq operator has given you:

the HTTPS URL of a running engine;
an engine-issued API key;
access to at least one configured and priced model.

Set these values in your server-side runtime:

export DEEPLINQ_BASE_URL="https://api.example.com"
export DEEPLINQ_API_KEY="sk_dl_..."

1. List available models

The models response is already filtered to the organization represented by your credential.

curl "$DEEPLINQ_BASE_URL/v1/models" \
  -H "Authorization: Bearer $DEEPLINQ_API_KEY"

2. Send a chat completion

Use a model returned by /v1/models, or use auto when your organization has enabled intent-based routing.

curl "$DEEPLINQ_BASE_URL/v1/chat/completions" \
  -H "Authorization: Bearer $DEEPLINQ_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "auto",
    "messages": [
      {
        "role": "user",
        "content": "Give me a concise explanation of hybrid retrieval."
      }
    ]
  }'

The response uses the OpenAI chat-completions shape, so existing SDK response handling continues to work.

3. Use an SDK

Install the OpenAI SDK in your server project:

bun add openai

server.ts

import OpenAI from "openai"

const client = new OpenAI({
  apiKey: process.env.DEEPLINQ_API_KEY,
  baseURL: `${process.env.DEEPLINQ_BASE_URL}/v1`,
})

const completion = await client.chat.completions.create({
  model: "auto",
  messages: [
    {
      role: "user",
      content: "Give me a concise explanation of hybrid retrieval.",
    },
  ],
})

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

Next steps

Authentication

Select between API keys and tenant-signed JWTs.

OpenAI SDK guide

Add streaming and per-user attribution.

Platform concepts

Understand the request and governance path.