OpenAI Response API and code interpreter tool

Question

OpenAI Response API and code interpreter tool

Siry, Gaetan 280

Hello,

I recently switched to the OpenAI Response API in Aure ( instead of the Assistants API, which is set to be deprecated soon )
I am running into issues with the code interpreter tool, where it works for a bit then stops working saying that it cannot access my file.

my container type is set to auto which is supposed to automatically manage the creation of environments.

Could it be that my container is expiring ?
once it expires how can i recover from it so that it does not impact accessing files ?

I do not see a lot of people creating their own container .. instead every one seems to be using auto

Pilladi Padma Sai Manisha 85 Reputation points Microsoft External Staff Moderator

2025-10-21T05:20:44.4766667+00:00
Hi @Siry, Gaetan,

It looks like you're having some trouble with the Code Interpreter in the OpenAI Response API within Azure that you are not being able to access your file could indeed relate to the container's timeout or expiration. Here's a quick breakdown of what you can do:

The Code Interpreter tool requires a container, and if it's set to "auto," it will manage its environment for you. However, these containers can expire if they haven’t been accessed within 20 minutes of being created. If your container is expiring, that's likely causing the issues you're facing.

What you should do:

You cannot “recover” an expired container; you must let the system create a fresh container by starting a new session or API call, don’t reuse the previous_response_id.

Any files attached to the old container are lost. Re-upload files with every new request that needs them.

“Auto” tries to reuse existing containers but cannot restore expired ones. If your workflow needs persistent file access, design it so files are re-uploaded or created anew each time.

This is normal most developers use “auto” because manual container management is rarely needed.

If your app needs long-running sessions, try to avoid long idle gaps and re-upload files when starting new sessions. For critical data, save results elsewhere before the container expires.

For more detailed information about how to manage containers and file access, you can refer to these resources:

About Containers in Code Interpreter

Hope this helps! If you have further questions or details, feel free to share!

2 answers

Your answer

Pilladi Padma Sai Manisha 85 Reputation points Microsoft External Staff Moderator

2025-10-21T05:20:44.4766667+00:00

Hi @Siry, Gaetan,

It looks like you're having some trouble with the Code Interpreter in the OpenAI Response API within Azure that you are not being able to access your file could indeed relate to the container's timeout or expiration. Here's a quick breakdown of what you can do:

The Code Interpreter tool requires a container, and if it's set to "auto," it will manage its environment for you. However, these containers can expire if they haven’t been accessed within 20 minutes of being created. If your container is expiring, that's likely causing the issues you're facing.

What you should do:

You cannot “recover” an expired container; you must let the system create a fresh container by starting a new session or API call, don’t reuse the previous_response_id.

Any files attached to the old container are lost. Re-upload files with every new request that needs them.

“Auto” tries to reuse existing containers but cannot restore expired ones. If your workflow needs persistent file access, design it so files are re-uploaded or created anew each time.

This is normal most developers use “auto” because manual container management is rarely needed.

If your app needs long-running sessions, try to avoid long idle gaps and re-upload files when starting new sessions. For critical data, save results elsewhere before the container expires.

For more detailed information about how to manage containers and file access, you can refer to these resources:

About Containers in Code Interpreter

Hope this helps! If you have further questions or details, feel free to share!

Answer 1

Hi Siry, Gaetan,
Sorry for the wait! Here is the answer to your question:

When I say “files are re-uploaded or created anew each time,” I mean that the code interpreter (when run in a new container) cannot see files that were only present inside the previous temporary environment, it must be given access again to any files you wish to use.

If you already uploaded files to Azure OpenAI Files (using their REST API or portal), you do not have to upload the original file again. Instead, you simply need to reference the correct file's file_id in your new request, this tells Azure to “mount” that file in the new container, making it accessible to code interpreter.

To ensure a new container (after expiration) can access your files:

Upload files once using the /openai/files endpoint, and retrieve the file_id for each file.

For each new API call that needs file access, include these file_id under the container.file_ids property. This way, any new container will automatically have access to your desired files.

You don’t have to manually re-upload the file every time, just use the file_id you already have.

POST https://YOUR-RESOURCE-NAME.openai.azure.com/openai/responses?api-version=2025-04-01-preview
Content-Type: application/json
api-key: YOUR_AZURE_OPENAI_KEY
{
  "input": [
    {
      "role": "user",
      "content": "analyze the attached data file"
    }
  ],
  "tools": [
    {
      "type": "code_interpreter",
      "container": {
        "type": "auto",
        "file_ids": [
          "file-xxxxxxxxxxxxxxxxxxxxxxxx"
        ]
      }
    }
  ],
  "model": "gpt-4.1",
  "stream": false
}

Replace "file-xxxxxxxxxxxxxxxxxxxxxxxx" with your actual Azure OpenAI file ID.
The same file ID can be used for any number of new requests, including after containers expire, as long as the file remains in Azure OpenAI Files storage.

I hope this helps clarify the situation! If you have more specific details or additional questions, feel free to share, and I can assist further!

could you please accept the answer!

Thankyou.

Answer 2

Siry, Gaetan 280

thank you .

My files are in Azure OpenAI files .

What exactly do you mean by "files are re-uploaded or created a new each time" ? even when the container expires, my files are still in Azure OpenAI files.

What is the process of re-uploading so that a new container would see the files that I pass using the file id ?

Can you share a sample REST request ?

thank you

Pilladi Padma Sai Manisha 85 Reputation points Microsoft External Staff Moderator

2025-10-23T06:24:59.7133333+00:00

Hi @Siry, Gaetan,
Did you get any chance to review the above response. Thank you!
Pilladi Padma Sai Manisha 85 Reputation points Microsoft External Staff Moderator

2025-10-24T02:05:32.24+00:00

Hi @Siry, Gaetan,
We haven’t heard from you on the last response and was just checking back to see if you have a resolution yet. In case if you have any resolution please do share that same with the community as it can be helpful to others. Otherwise, will respond with more details and we will try to help.
Thankyou.
Siry, Gaetan 280 Reputation points

2025-10-24T02:08:38.2433333+00:00

All set thank you for your help
Pilladi Padma Sai Manisha 85 Reputation points Microsoft External Staff Moderator

2025-10-24T05:32:55.0466667+00:00

Hi @Siry, Gaetan,
Kindly accept the answer if it solved your problem.
Thankyou!

Share via

OpenAI Response API and code interpreter tool

2 answers

Your answer