Problem

OpenClaw currently lacks native support for Azure OpenAI (Azure AI Foundry) as a model provider.

Why This Matters

• Enterprise adoption: Azure OpenAI is one of the most popular enterprise LLM deployment options, offering GPT-4o, GPT-4o-mini, and other models with enterprise-grade security and compliance
• Azure credits: Many users have Azure credits they want to leverage (Azure Sponsorship, Enterprise Agreements, MSDN subscriptions, etc.)
• Current workaround is cumbersome: Using Azure OpenAI today requires setting up a LiteLLM proxy as an intermediary, which adds deployment complexity, operational overhead, and latency

Technical Details

The Azure OpenAI API differs from the standard OpenAI API:

• Uses api-key header instead of Authorization: Bearer
• Different endpoint format: https://<resource>.openai.azure.com/openai/deployments/<deployment>/chat/completions?api-version=<version>
• Requires explicit deployment names rather than model names
• Requires API version parameter

Proposed Solution

Add azure or azure-openai as a native provider option, similar to the existing openai, anthropic, and google providers.

Example Configuration

{
  "models": {
    "providers": {
      "azure": {
        "endpoint": "https://<resource>.openai.azure.com",
        "apiKey": "...",
        "apiVersion": "2024-10-21",
        "deployments": {
          "gpt-4o-mini": "my-gpt4o-mini-deployment"
        }
      }
    }
  }
}

Benefits

• Direct Azure integration without additional infrastructure
• Simpler configuration for Azure users
• No extra proxy layer needed (reduced latency and complexity)
• Better alignment with enterprise deployment patterns
• Enables users to leverage existing Azure investments and credits

Additional Context

Azure OpenAI Service documentation: https://learn.microsoft.com/en-us/azure/ai-services/openai/