I’m building a chat based agent using gpt-5 and the openai agents sdk. We stream responses to the user. The agent commonly uses a few tools and sometimes does a few batches of tool calls. I’m trying to utilize gpt-5’s pre-disposition to provide tool preambles - user facing explanations of what it w…

[image] matt.ambrogi: Update: So really there’s no “problem” to be fixed here. But GPT-5’s eagerness to provide the user with updates about what it is going to do before it does it has interested UX impacts Most users of agentic applications have become accustomed to seeing reasoning tokens st…

The AI hasn’t emitted a tool call when it is writing to the user still. It is natural language “content” of a turn that doesn’t need to be held out or used differently. The only part of “decided” that is definite is when the AI writes, “Let me check my weather tool for Walla Walla for you now. (sp…

I understand all of the above. But my point is that if you as a developer desired to display this pre tool call statement of intent differently, there is no way to do so. This is a weird UX. In every agentic application I use, the preamble or reasoning is displayed differently than final response t…

[image] [image] It is typical for the AI to simply call tools immediately, in this case, without having any idea what it would be picking from randomly, so the reasoning doesn’t quite reveal if that actually got done. The trigger that could be immediately parsed (or held back for inspecti…

I think you are misinterpreting my point. If I query gpt-5 with access to tools there are two thing it might do before calling a tool. The internal reasoning which is what I believe you’ve displayed above. This is identifiable as a reasoning item in the api response. You can get the text by inclu…

A “preamble” isn’t a real thing nor introduced with GPT-5. It has always been possible for the AI to write language “content” and also then call a tool (the order is never flipped, because emitting a tool means awaiting the response in a follow-up API call). The AI model generates tokens one-at-a-t…

“A “preamble” isn’t a real thing nor introduced with GPT-5.” Regarding this I know what you mean but its not quite the full picture. Prior models were trained to heavily prefer jumping right into tool calling. So a preamble was possible but very unlikely. GPT-5 has been trained to provide these more…

Source on others having issues with model adherence to not providing preambles GPT-5 mini not following instructions for preambles opened 05:34PM - 18 Aug 25 UTC (UTC) [image] MSLaFaver-PASS …

It’s a clunker of a model anyway. Look at this, can’t even tell that “developer” is not the user themselves, making it garbage for closed-domain API product development: [image] “Per my instructions” indeed. Talking about the developer role as if it were from a user. Because a developer is no mor…

Yea the model is very interesting. Love it in some respects. Hard to work with in others. Thanks for your input.

How differentiate preamble from final response during streaming?

API

matt.ambrogi August 22, 2025, 6:52pm 4

I would happily just get rid of the preambles and display the reasoning summary instead. But per this thread I’ve found the reasoning summary before tool calls is often empty.

Topic		Replies	Views
Realtime API Preamble Inconsistent API realtime , gpt-realtime	1	314	October 11, 2025
More consistent tool calling for GPT-5 API gpt-5	7	1711	October 7, 2025
Auto tool call streaming differentiation is unintuitive Feedback api	3	291	May 19, 2025
How to Prevent Tool Output Summarization in Responses API API tool , gpt-41	27	580	May 26, 2025
How to stop GPT-5 from exposing reasoning before tool calls? Prompting chatgpt , gpt-5	1	1595	September 25, 2025

How differentiate preamble from final response during streaming?

Related topics