Streaming the final output

[taha-yassine]

Hi,

I was wondering if there’s a way to stream the final output only, rather than every token in an agent run. I couldn’t find anything about this in the documentation.

I realize this might not be possible conceptually, since the final output is only determined once the model finishes generating. But I wanted to check if there’s any workaround to make it somehow work.

Thanks!

[seratch]

Hi, thanks for asking the question. You can filter streaming events by their types this way: https://github.com/openai/openai-agents-python/blob/main/examples/basic/stream_text.py

[taha-yassine]

Hi, thanks for asking the question. You can filter streaming events by their types this way: https://github.com/openai/openai-agents-python/blob/main/examples/basic/stream_text.py

Thanks you for your answer. However, this includes every token generated, not the final output only. What I would like to be able to do is to stream the final output only.

[seratch]

Unfortunately, it is not possible. Thanks again for this question!