How to use fine-tuned LLMs from an account but billing for another account

starfriend 0 Reputation points
2025-05-09T17:05:30.54+00:00

Hi there,

I have completed GPT finetuning (Azure openAI) and got fine-tuned LLM models though one my Azure account but which reaches its expense limit, so I want to use my another Azure account to continue work on these fine-tuned models (e.g., evaluation, continuous FT, ...) and the new expense will be only accounted in my second account, is this possible? How can I do this?

I see there is a way to transfer models, will that solve my issue, and will the transferring cause new expense in old account?

Or if there are any other approaches?

Thank you!

Azure OpenAI Service
Azure OpenAI Service
An Azure service that provides access to OpenAI’s GPT-3 models with enterprise capabilities.
3,974 questions
0 comments No comments
{count} votes

1 answer

Sort by: Most helpful
  1. Pavankumar Purilla 7,100 Reputation points Microsoft External Staff Moderator
    2025-05-09T22:01:31.5466667+00:00

    Hi starfriend,

    You can move your fine-tuned Azure OpenAI models to another Azure account to ensure that all future usage and fine-tuning costs are billed to the second account. To do this, follow these steps:

    Export the fine-tuned model from the original Azure OpenAI resource:

    Go to the Azure portal → Azure OpenAI resource → Fine-tuned models.

      Select the fine-tuned model you want to transfer.
      
         Use the **Export** option to download the model archive.
         
         **Import the model into the new Azure account**:
         
            Sign in to your second Azure account.
            
               Ensure you have an Azure OpenAI resource set up with access to the same base model (e.g., `gpt-35-turbo` or `gpt-4`).
               
                  Go to the Azure portal → Azure OpenAI resource → Fine-tuned models → Import, and upload the exported model archive.
                  
                  **Update your application configuration**:
                  
                     Point your applications or API clients to the new resource endpoint and use the API key from the second Azure account.
                     
                     **Continue fine-tuning or evaluation in the new account**:
                     
                        Any further fine-tuning, evaluation, or inference done under the new resource will be billed to the second account.
                        
    

    The model transfer itself does not generate significant additional cost on the old account beyond the export operation. Note that simply sharing API keys or using service principals between accounts will not change where the costs are billed — the billing always follows the resource’s subscription.

    I hope this information helps.


Your answer

Answers can be marked as Accepted Answers by the question author, which helps users to know the answer solved the author's problem.