When ChatGPT tries to send a request to the plugin, it sometimes gets cut off while generating the message. This happens when the request message reaches approximately 2500 characters, causing ChatGPT to be unable to complete and send the request.
How can i increase the request message limit?
Is it really character limitation?
Maybe remove the vowels from the answer?
I have attached an example output. I get similar output like this.
It stops after reaching certain character limit.
I want chatGPT to complete till the end.
You could try to shorten the keys. Change the answer to this.
{“sS”:… “sS”:…
And add this to the openapi definition as description
strokeSharpness:sS
Can you please provide me some examples on “openapi definition as description” on shorten keys?
May be a link is fine.
This would be useful to me.
Thanks
Sorry, I didn’t publish that anywhere.
So this is the first time I’ve posted that.
You can take your current openapi definition, paste it into ChatGPT together with what I wrote. I think it can build that for you.
Thank you. I will try that. Thanks for your prompt response.
You are welcome.
Would be nice if you could tell us when it works.
It worked by allowing 2-3 times more text response.
Is there any possibility / ideas to increase that further?
Did you do that shortening of the keys to all keys or just the strokeSharpness:sS?
I mean there is versionNonce:vN, boundElementIds:BEI, version:v, false:f, center:cnt, isDeleted:iD… etc…
You can take your output together with following prompt to ChatGPT:
Hey Assistant,
help me shorten this:
{
"output":...
}
e.g.
strokeSharpness:sS, boundElementIds:bEI,…
Create a python code that replaces the longer keys and values like false:f before I output it in my API
Here is my current API code where the json is created:
[your code]
also here is my openapi definition:
[your definition]
change that accordingly

You wouldn’t believe how in banking we’ve tried to keep the answers of APIs as short as possible to get the least amount of data for transmission.