OpenAI Assistants keep referencing the documents used

Here is what my assistants usually return:

"Unlock the door to your future with Studykai’s study abroad programs in Spain! :earth_americas::mortar_board: Dive into a world of affordable, career-focused education that blends academic rigor with real-world business insights. Highlights include:

Affordable Education: Enjoy lower living costs compared to other European countries, spending about €900 to €1,200 monthly【4:0†source】.
Cultural Immersion: Experience Spain’s vibrant culture, beautiful landscapes, and rich history【4:0†source】.
Career Advantages: AIFS reports 58% of students gain their first job due to skills from study abroad experiences【4:2†source】.
Modern Facilities: Benefit from state-of-the-art libraries and science labs at top universities like the Polytechnic University of Valencia【4:0†source】.
Flexibility: Programs range from 2 to 10 weeks, incorporating internships and housing options【4:13†source】.

Join us in Barcelona or Madrid, where learning meets leisure. Your global career starts here. #StudyAbroad #Education #CareerOpportunities #Spain #GlobalExperience #StudentLife :star2:"

  1. How do I make the assistants to stop referencing the docs and add characters like: [x:0†source]
  2. Is it possible to adjust them within make.com or the openai playground?

Currently, after each assistant result, I have another chat completion module that adjusts the text. But it is not always correct plus it uses additional costs.

I’ve been searching all over the place for a solution, so help is much appreciated =))


blueprint (5).json (206.0 KB)

solved it

You have to use the text parser “Replace” module after each openai module and use the following pattern

Hope it helps anyone :ok_hand:

1 Like