Impossible to work with ChatGPT at this point

ChatGPT
,
1

I’m having several constant issues with ChatGPT:

  • After about 2 minutes, responses show “Streaming interrupted. Waiting for the full message.” I can’t even remember the last time I got a response to my prompt without seeing this
  • After around 5–6 minutes they always end with “Error in request. Please try again.” Two months ago ChatGPT could work for like 45-60 minutes. One month ago it was down to 22-25 minutes. Now I have it EVERY 4-6 minutes.
  • “This chat has become too long” appears after only a few messages, even when there’s barely any actual content. In the last 3 days I needed to create 8 new chats because every chat was too long way way way way WAY WAY faster than two or even one month ago.
  • Model 6 keeps activating itself automatically, no matter which model I manually select.

At this point it’s 100% impossible to do anything with ChatGPT, I really don’t get why OpenAI crippled the best AI down to this. Any explanation for this? A month ago I was barely able to continue my project but it gets worse and worse every single day.

Is anyone else experiencing this?

2

You can set this in config.toml
model = “gpt-5.6-sol”
model_reasoning_effort = “medium”
There is somewhere you also set something that prevents the model upgrading the effort, but I don’t remember where that is. Sol is working well

3

I’m already using 5.6 (had to re-change the model settings manually), all those problems are while using 5.6 mid to high. The project workflow was flawless with 5.6 two months ago and it gets crippled down every day since then. I’m at a point where I can’t do any progress at all.

@OpenAI_Support Any word on this? Anything?