Provider integrations

Send x-orq-integration-<provider> to choose which connected integration the AI Gateway uses for that provider. The value is the integration id or its exact display name. A model that already pins an integration keeps that integration.

Overview

Each provider can have more than one integration in a workspace. Without this header, a request uses the integration id on the request when one is set, then the provider’s default integration. The header applies when the model does not pin an integration and the request does not include a provider credential of its own. The <provider> segment is the integration identifier, such as anthropic, openai, google, or google-ai. It matches the provider on the model being called. A header for one provider does not change the integration used for another provider in the same request, including fallbacks. Names are unique for a provider inside a workspace. Two integrations for the same provider cannot share a display name.

Example

curl -X POST https://my.orq.ai/v3/router/chat/completions \
  -H "Authorization: Bearer $ORQ_API_KEY" \
  -H "Content-Type: application/json" \
  -H "x-orq-integration-openai: Production" \
  -d '{
    "model": "openai/gpt-4o",
    "messages": [{ "role": "user", "content": "Summarize the repository." }]
  }'

import OpenAI from "openai";

const client = new OpenAI({
  apiKey: process.env.ORQ_API_KEY,
  baseURL: "https://my.orq.ai/v3/router",
  defaultHeaders: {
    "x-orq-integration-openai": "Production",
  },
});

const response = await client.chat.completions.create({
  model: "openai/gpt-4o",
  messages: [{ role: "user", content: "Summarize the repository." }],
});

import os
from openai import OpenAI

client = OpenAI(
    api_key=os.environ["ORQ_API_KEY"],
    base_url="https://my.orq.ai/v3/router",
    default_headers={"x-orq-integration-openai": "Production"},
)

response = client.chat.completions.create(
    model="openai/gpt-4o",
    messages=[{"role": "user", "content": "Summarize the repository."}],
)

Configuration

Header	Value	Description
`x-orq-integration-<provider>`	Integration id or exact display name	Selects the workspace integration for that provider.

One header per provider. x-orq-integration-anthropic and x-orq-integration-google can be sent on the same request. A header for a provider the request does not call is ignored.

Result	When
`400`	The header for a provider on the request is empty, or no integration in the workspace has that id or exact display name.
`409`	More than one integration for that provider has the same display name. Send the integration id.

Claude Code

Claude Code forwards extra headers through ANTHROPIC_CUSTOM_HEADERS. Keep x-orq-api-key when the session uses an Anthropic credential, and add the integration header on its own line:

Shell

export ANTHROPIC_BASE_URL="https://my.orq.ai/v3/anthropic"
export ANTHROPIC_MODEL="anthropic/claude-sonnet-4-6"
export ANTHROPIC_CUSTOM_HEADERS="x-orq-api-key: $ORQ_API_KEY"$'\n'"x-orq-integration-anthropic: Production"

When the Orq.ai key is the Anthropic client credential, the integration header is enough:

Shell

export ANTHROPIC_BASE_URL="https://my.orq.ai/v3/anthropic"
export ANTHROPIC_AUTH_TOKEN="$ORQ_API_KEY"
export ANTHROPIC_API_KEY=""
export ANTHROPIC_MODEL="anthropic/claude-sonnet-4-6"
export ANTHROPIC_CUSTOM_HEADERS="x-orq-integration-anthropic: Production"

An Anthropic API key or Claude subscription credential on the request is still the credential for that call. The header selects a workspace integration when the request uses the workspace’s Anthropic integration.

generateContent

Bare Gemini model ids on https://my.orq.ai/v3/google resolve as google/<model>, so the header is x-orq-integration-google. Google AI Studio integrations use the identifier google-ai and the header x-orq-integration-google-ai when the model provider is google-ai.

import { GoogleGenAI } from "@google/genai";

const client = new GoogleGenAI({
  apiKey: process.env.ORQ_API_KEY,
  httpOptions: {
    baseUrl: "https://my.orq.ai/v3/google",
    headers: {
      "x-orq-integration-google": "Production",
    },
  },
});

const response = await client.models.generateContent({
  model: "gemini-2.5-flash",
  contents: "Explain the repository structure.",
});

import os
from google import genai
from google.genai import types

client = genai.Client(
    api_key=os.environ["ORQ_API_KEY"],
    http_options=types.HttpOptions(
        base_url="https://my.orq.ai/v3/google",
        headers={"x-orq-integration-google": "Production"},
    ),
)

response = client.models.generate_content(
    model="gemini-2.5-flash",
    contents="Explain the repository structure.",
)

Limitations

Limitation	Impact	Workaround
A model pins an integration	The header for that provider is ignored for that model	Change the integration on the model, or call a model that does not pin one
The request includes a provider credential	That credential is used for the call	Omit the provider credential to select a workspace integration
Display names must match exactly	`Production` does not match `production`	Send the integration id, or the display name as stored

Overview

APIs

MCP Portal

Routing

Security & Privacy

Model Capabilities

Server Tools

Common Scenarios

Overview

Example

Configuration

Claude Code

generateContent

Limitations

​Overview

​Example

​Configuration

​Claude Code

​generateContent

​Limitations

Overview

Example

Configuration

Claude Code

generateContent

Limitations