Enable Azure Foundry deployments of claude sonnet to be used with holmesgpt

[ZOBHeile]

What would you like to be added?

Hi,

currently I can only configure an anthropic API Key and not the endpoint. Azure Foundry gives the ability in enterprise environments to use Claude Sonnet. For this to work, we need to be able to configure the endpoint of a self-deployed model.

Example Curl:

curl -X POST "https://XXXXXX.services.ai.azure.com/anthropic/v1/messages"
-H "Content-Type: application/json"
-H "x-api-key: YOUR_API_KEY"
-H "anthropic-version: 2023-06-01"
-d '{
"max_tokens": 1000,
"temperature": 0.7,
"system": "You are a helpful assistant.",
"messages": [
{ "role": "user", "content": "What are 3 things to visit in Seattle?" }
]
"model": "claude-sonnet-4-6"
}'

Why is this needed?

Enable use of Azure AI Foundry hosted models.

[arikalon1]

Hey @ZOBHeile

Can you try setting azure model this way:

modelList:
  azure-sonnet-4-6:
    model: anthropic/claude-sonnet-4-5
    api_key: "{{ env.AZURE_API_KEY }}"
    api_base: https://your-resource.services.ai.azure.com/anthropic
    temperature: 1

Does that work?

[ZOBHeile]

@arikalon1 thanks for support, I use the CLI. Which environment variables do I need to set?

[arikalon1]

Hey @ZOBHeile

Please use: TEMPERATURE, API_BASE, API_KEY

Does that help?