Unable to use Azure OpenAI provider

[lvnr]

The example in docs doesn't show an example for Azure OpenAI which requires a few more params to be set to run:
i've tries adding azureResourceName, azureDeploymentId, etc to the config like this:

import { createPortkey } from '@portkey-ai/vercel-provider';
import { streamText } from 'ai';

const portkeyConfig = {
      "provider": "openai", // Choose your provider (e.g., 'anthropic')
      "api_key": "OPENAI_API_KEY",
      "override_params": {
        "model": "gpt-4o" // Select from 250+ models
  }
};

const portkey = createPortkey({
  apiKey: 'YOUR_PORTKEY_API_KEY',
  azureResourceName: process.env.AZURE_OPENAI_RESOURCE_NAME,
  azureDeploymentId: process.env.AZURE_OPENAI_DEPLOYMENT_ID,
  azureApiVersion: process.env.AZURE_OPENAI_API_VERSION,
  azureModelName: model,
  Authorization: `Bearer ${process.env.AZURE_OPENAI_API_KEY}`,
  config: portkeyConfig,
});

Now this is what I see in Portkey Logs:

{
  "url": "https://undefined.openai.azure.com/openai/deployments/undefined/chat/completions?api-version=undefined",
  "method": "POST",
  "headers": {
    "accept": "*/*",
    "accept-encoding": "gzip, br",
    "accept-language": "*",
    "cf-connecting-ip": "5.77.203.18",
    "cf-ipcountry": "AM",
    "cf-ray": "9390224c4f037d29",
    "cf-visitor": "{\"scheme\":\"https\"}",
    "connection": "Keep-Alive",
    "content-length": "3085",
    "content-type": "application/json",
    "host": "api.portkey.ai",
    "sec-fetch-mode": "cors",
    "user-agent": "node",
    "x-forwarded-proto": "https",
    "x-portkey-strict-open-ai-compliance": "false",
    "x-real-ip": "5.77.203.18"
  },
  "body": {
    ...
  },
  "portkeyHeaders": {
    "x-portkey-api-key": "******",
    "x-portkey-azure-api-version": "2025-01-01-preview",
    "x-portkey-azure-deployment-id": "gpt-4o-mini",
    "x-portkey-azure-endpoint-name": "https://denzo-dev.openai.azure.com/openai/deployments/gpt-4o-mini/chat/completions?api-version=2025-01-01-preview",
    "x-portkey-azure-model-name": "gpt-4o-mini",
    "x-portkey-azure-resource-name": "denzo-dev",
    "x-portkey-config": "{\"provider\":\"azure-openai\",\"model\":\"gpt-4o-mini\",\"api_key\":\"********\",\"override_params\":{\"model\":\"gpt-4o-mini\",\"stream_options\":{\"include_usage\":true}},\"metadata\":{\"_user_id\":\"\",\"chat_id\":\"\",\"environment\":\"development\"}}",
    "x-portkey-package-version": "portkey-1.7.2",
    "x-portkey-runtime": "node",
    "x-portkey-runtime-version": "v18.18.2",
    "x-portkey-trace-id": "..."
  }
}

Note that the URL has undefined in all parametrized places, even though portkeyHeaders seems to have all the params picked up.

Not sure if i'm doing something wrong or if it's a bug in the lib.

The endpoint is working fine, since I tried using it without the vercel provider lib and with a virtual key, which worked ok.

[narengogi]

@lvnr strange, you're sending the correct headers as far as I can tell,
couple of things I can see in your code snippet

const portkeyConfig = {
      "provider": "openai", // Choose your provider (e.g., 'anthropic')
      "api_key": "OPENAI_API_KEY",
      "override_params": {
        "model": "gpt-4o" // Select from 250+ models
  }
};

1. here, provider should be set to "azure-openai" (not required if you're using a virtual key)
2. I see that you're passing a config as well, does it have a virtual key in it? and are the resourceName and deploymentId not set in the virtual key? because the values from the virtual key will override the ones you've passed in the headers