How to connect Azure OpenAI to MultimodelChat
Connect Azure OpenAI deployments to your BYOK chat workspace
Azure OpenAI serves OpenAI models from your own Azure resource. Create the resource, deploy a model, then connect the endpoint and key to MultimodelChat as an OpenAI-Compatible provider.
VariesAzure OpenAI is billed through your Azure subscription. Access and model availability may require an application or quota approval for your resource.
How to connect Azure OpenAI
Create an Azure OpenAI resource in the Azure portal, deploy a model, and copy the resource endpoint and key from the Keys and Endpoint page. Then connect it in MultimodelChat as an OpenAI-Compatible provider using your endpoint URL and deployment name.
- Create an Azure OpenAI resource (request access if prompted)
- Deploy a model and note the deployment name
- Copy the endpoint URL and a key
- Connect as OpenAI-Compatible in MultimodelChat
Connect it to MultimodelChat
Once you have the required credentials, start your free trial, add this provider in Settings, and save the connection. MultimodelChat encrypts the API key before storing it.
Available models on Multimodel Chat
Azure serves whatever models you deploy in your resource. In MultimodelChat, use the deployment name you chose in Azure as the model ID.
Step-by-step guide
Create an Azure OpenAI resource
Provision the resource in the Azure portal.
- Sign in to https://portal.azure.com
- Create a new Azure OpenAI (or Microsoft Foundry) resource
- Choose your subscription, region, and pricing tier
- If access is gated, submit the requested access form and wait for approval
Deploy a model
Deployments are how Azure serves models.
- Open the resource in the Azure AI Foundry / portal view
- Go to Deployments and create a deployment
- Choose a base model and give the deployment a name
- Remember the deployment name — it is the model ID you use in the app
Copy the endpoint and key
Find these on the resource's Keys and Endpoint page.
- Open Keys and Endpoint for your resource
- Copy the endpoint URL, e.g. https://your-resource.openai.azure.com
- Copy KEY 1 (or KEY 2)
- Treat the key as a secret — it is not shown again in the app
Azure endpoints are per-resource. Do not paste another resource's key.
Connect it as an OpenAI-Compatible provider
Azure exposes an OpenAI-compatible surface.
- In MultimodelChat, go to Settings → API Keys
- Select "OpenAI-Compatible"
- Base URL: your Azure endpoint (for example https://your-resource.openai.azure.com)
- Paste your Azure key
- Model ID: your Azure deployment name
- Save — MultimodelChat tests the connection
Azure authentication and URL paths can differ between the classic and OpenAI v1 surfaces. If the test fails, check the current Microsoft Learn guidance for your resource type.
Common issues
Swipe horizontally to see the full table.
| Issue | Fix |
|---|---|
| Model not found | Use your Azure deployment name as the model ID, not the base model name |
| Authentication failed | Confirm you copied KEY 1 from the same resource as the endpoint |
| Resource access denied | Some subscriptions need an access application approved before the resource works |
| Wrong endpoint path | Check Microsoft Learn for the current OpenAI-compatible URL path for your resource type |
Frequently asked questions
Why does Azure use a deployment name?
In Azure you deploy a base model under your own name. Requests are addressed to that deployment, so the deployment name is what you enter as the model ID.
Is Azure OpenAI the same as OpenAI?
No. It is OpenAI's models hosted inside your Azure subscription, with its own endpoints, keys, quota, and billing through Azure.
Set up another provider
Related guides
Provider requirements and pricing last reviewed October 10, 2026. Check the linked provider documentation for current terms before purchasing credits.
Ready to connect your Azure OpenAI key?
Paste your API key and Multimodel Chat tests it live against the provider before saving. The key is encrypted in this browser's vault and used only for your requests.
Start the 7-day free trialNew to the workflow? See how it works in three steps