I mean… What else do I need to say about this 
This new UI push has been extremely annoying to work with. I honestly wouldn’t be using these frameworks were it not being shoved down my throat by my boss.
Regardless. Fix your stuff LangSmith team. Try having a staging environment run through a little QA testing.
Hello @UsamaBinMuzaffar ,
Sorry for the frustration you are experiencing. The issue is that 'minimal' is not a valid value for reasoning.effort with this model as stated in the official OpenAI GPT-5.4 model page.
Supported values are 'none', 'low', 'medium', 'high', and 'xhigh' for gpt-5.4-mini, and the minimal one shown in the playground dropdown is incorrect (I was able to replicate it on my end as well).
While the fix gets patched, can I recommend that you use the low which is equivalent to minimal in case of gpt-5-4-mini.
I would also recommend that you open a ticket for this here
Also Tagging official Langchain team member so they can also get notified: @wfh
Hi @UsamaBinMuzaffar - thanks for your feedback and flagging this issue. Just to confirm did you have “minimal” selected when that exception was triggered? If so, you’ll want to select “Default” as @keenborder786 mentioned (thank you)
That being said we’ll get a patch out ASAP to make sure that minimal isn’t shown for OpenAI models 5.1 and onward, additionally we’re working towards standardizing reasoning efforts across models which should avoid issues like this in the future.
Thank you again and appreciate your patience here. If you have any other feedback, questions or concerns please do let us know.
