> ## Documentation Index
> Fetch the complete documentation index at: https://docs.orq.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Base URLs

> Point the Orq.ai SDKs, AI Gateway clients, and OpenTelemetry exporters at a self-hosted or on-premise deployment by overriding the default base URL with server_url, serverURL, base_url, or ORQ_BASE_URL.

Every **Orq.ai** client targets the managed cloud at `https://my.orq.ai` unless told otherwise. Self-hosted and on-premise deployments serve the same API paths under their own hostname, so every client must be pointed at that hostname explicitly. No client reads the base URL from the environment on its own.

## Base URLs

| Client                       | Setting                       | Managed cloud                 |
| ---------------------------- | ----------------------------- | ----------------------------- |
| Python SDK (`orq-ai-sdk`)    | `server_url`                  | `https://my.orq.ai`           |
| Node.js SDK (`@orq-ai/node`) | `serverURL`                   | `https://my.orq.ai`           |
| OpenAI-compatible clients    | `base_url` / `baseURL`        | `https://my.orq.ai/v3/router` |
| OpenTelemetry exporter       | `OTEL_EXPORTER_OTLP_ENDPOINT` | `https://my.orq.ai/v2/otel`   |

Only the host changes between deployments. The paths stay the same: a deployment reachable at `https://orq.example.com` serves the router at `https://orq.example.com/v3/router` and the OTLP endpoint at `https://orq.example.com/v2/otel`.

Set `ORQ_BASE_URL` alongside `ORQ_API_KEY` and read both from the environment, so the same code runs against the managed cloud and a self-hosted deployment without edits:

```bash theme={"theme":{"light":"github-light","dark":"github-dark"}}
export ORQ_BASE_URL="https://orq.example.com"
export ORQ_API_KEY="<api-key>"
```

## Orq SDKs

Pass the base URL as `serverURL` (Node.js) or `server_url` (Python):

<CodeGroup>
  ```typescript TypeScript theme={"theme":{"light":"github-light","dark":"github-dark"}}
  import { Orq } from "@orq-ai/node";

  const client = new Orq({
    apiKey: process.env.ORQ_API_KEY,
    serverURL: process.env.ORQ_BASE_URL ?? "https://my.orq.ai",
  });
  ```

  ```python Python theme={"theme":{"light":"github-light","dark":"github-dark"}}
  import os

  from orq_ai_sdk import Orq

  client = Orq(
      api_key=os.environ["ORQ_API_KEY"],
      server_url=os.environ.get("ORQ_BASE_URL", "https://my.orq.ai"),
  )
  ```
</CodeGroup>

<Note>
  The Node.js option is `serverURL`, with `URL` in uppercase. `serverUrl` is not a valid option and leaves the client pointed at the managed cloud.
</Note>

## AI Gateway

OpenAI-compatible clients take the router path as their base URL:

<CodeGroup>
  ```bash cURL theme={"theme":{"light":"github-light","dark":"github-dark"}}
  curl -X POST "$ORQ_BASE_URL/v3/router/chat/completions" \
    -H "Authorization: Bearer $ORQ_API_KEY" \
    -H "Content-Type: application/json" \
    -d '{...}'
  ```

  ```typescript TypeScript theme={"theme":{"light":"github-light","dark":"github-dark"}}
  import OpenAI from "openai";

  const client = new OpenAI({
    apiKey: process.env.ORQ_API_KEY,
    baseURL: `${process.env.ORQ_BASE_URL ?? "https://my.orq.ai"}/v3/router`,
  });
  ```

  ```python Python theme={"theme":{"light":"github-light","dark":"github-dark"}}
  import os

  from openai import OpenAI

  client = OpenAI(
      api_key=os.environ["ORQ_API_KEY"],
      base_url=f"{os.environ.get('ORQ_BASE_URL', 'https://my.orq.ai')}/v3/router",
  )
  ```
</CodeGroup>

The same substitution applies to every tool that accepts a base URL, including the coding agents, chat interfaces, and frameworks under [Integrations](/docs/ai-studio/integrations/overview). Those pages document the managed cloud host, so replace `https://my.orq.ai` with the deployment hostname when configuring them.

## OpenTelemetry

Exporters read the endpoint from the environment. Point it at the deployment host:

```bash theme={"theme":{"light":"github-light","dark":"github-dark"}}
export OTEL_EXPORTER_OTLP_ENDPOINT="$ORQ_BASE_URL/v2/otel"
export OTEL_EXPORTER_OTLP_HEADERS="Authorization=Bearer $ORQ_API_KEY"
```

Exporters that require the full signal path use `$ORQ_BASE_URL/v2/otel/v1/traces`. See the [Observability quickstart](/docs/ai-studio/observability/quickstart) for the full exporter setup.

## Troubleshooting

**Requests reach the managed cloud instead of the deployment.** The base URL was not applied. In Node.js, confirm the option is spelled `serverURL` and not `serverUrl`. In Python, confirm `server_url` is passed to the `Orq` constructor rather than to the individual method call. When the option is missing or misspelled the SDKs fall back to the host published in the OpenAPI specification, `https://api.orq.ai`, so calls succeed against the managed cloud instead of failing outright.

<Card title="On-Prem Requirements" icon="server" href="/docs/enterprise/on-prem-requirements">
  Infrastructure requirements for deploying **Orq.ai** on-premise, including the DNS record and TLS certificate for the platform hostname.
</Card>