AI completion with extra quotation mark

Problem description

When every I type one quotation mark and wait for AI completion, the is going to be an extra quotation mark.

Expected behavior

Case 1:
When I am entering the key of a dictionary, I type "
The completion should be “keywords”: keywords,
Case 2:
I type print("
The completion should be something like print(“Fetch news finished!”)

Actual behavior

Case 1:
The actual completion is “keywords”: self.keywords,"
Notice there is an extra quotation mark at the end, because the text editor first auto completes the quotation mark, and then AI completed the part “keywords”: self.keywords,
Case 2:
The actual completion is print(“Fetch news finished!”)"
The reason is the same as case 1.

Steps to reproduce

Description above is pretty clear.




Mac OS

Device if mobile



Core membership

Please upload screenshots

When you type one quotation mark, the editor (not the AI) automatically inserts a second quotation mark for you. Next, the AI then suggests content based on your quotation mark, which would lead to the above dilemma.

1 Like

:wave: Welcome @forfridden!

Well, yeah, AI cannot insert code after your cursor position. This isn’t really something that Replit can do anything about.


They should remove the dbl quote auto function since the ai exists. I find it harbours more errors.

  1. AI hallucinates, a lot.
  2. For people who don’t use the AI, the double quotes are quite useful.