Return usage when streaming chat completions.

[samshapley]

Confirm this is a feature request for the Python library and not the underlying OpenAI API.

• This is a feature request for the Python library

Describe the feature or improvement you're requesting

Is it possible to return token usage when streaming in the final chunk?

otherwise we need to also use TikToken in an application where both streaming is enabled vs disabled.

Additional context

No response

[rattrayalex]

Thanks for the suggestion!

This sounds like a feature request for the underlying OpenAI API and not the Python library, so I'm going to go ahead and close this issue.

Would you mind reposting at community.openai.com?

[jackmpcollins]

There is a community post for this here: https://community.openai.com/t/openai-api-get-usage-tokens-in-response-when-set-stream-true/141866
The corresponding issue on the JS client is openai/openai-node#506 and comments there indicate this will be added to the API.

@rattrayalex Would you consider reopening this issue so we can subscribe to be notified when streaming usage has been added to the python client? Thanks

This is a requested feature for the package I maintain jackmpcollins/magentic#74

[rattrayalex]

Sorry, I can't guarantee that we'd remember to close this issue when the feature ships, so I'd rather not reopen.

If we do remember, however, we'll leave a comment here so you'll be notified.

[samshapley]

For anyone looking in the future, this is now shipped

https://x.com/openaidevs/status/1787573348496773423?s=46&t=jm46NyFc_ht8JJU4ZFVivw