Who is gpt-4o agent successor in the gpt-5 serie?

API
,
1

Hello everyone,

For the past year I’ve been building agents using Azure OpenAI gpt-4o 2024-11-20.

I wanted to upgrade to one of the model of the gpt-5 serie to gain performance notably in SWE and Instruction following. My main use cases are text-to-SQL, python coding and RAG and I require:

  • Quick response: so very little to no reasoning
  • Tool call

I found in the OpenAI documentation that gpt-5-chat-latest is the natural successor of gpt-4o:
GPT-5 System Card | OpenAI

image
image757×472 16.7 KB

Using GPT-5 - OpenAI API

System card name API alias
gpt-5-thinking gpt-5
gpt-5-thinking-mini gpt-5-mini
gpt-5-thinking-nano gpt-5-nano
gpt-5-main gpt-5-chat-latest
gpt-5-main-mini [not available via API]
  • gpt-5-chat-latest doesn’t support tool call in OpenAI.
  • gpt-5-chat in Azure OpenAI does support tool call, but tends to not call them much and output everything in the chat (plus the model is in preview).
  • gpt-5 has a too long time to first token (~40s) for my use case even with reasoning_effort “low” and verbosity “low”

What are you thoughts on that?

1 Like
2
2 Likes
3

Try o4-mini.

It reasons, but reasons fast, then a rapid token spew across your screen.

It has better-quality understanding than gpt-5, and then you just have to argue about style.

Where’s our o4??