Dynamic Prompting in GPT Conversations

Fredric Cliver
3 min readOct 13, 2023


Dynamic prompting is a technique used to guide the behaviour of the GPT model in conversational setups. By interspersing system instructions throughout the conversation, developers can achieve finer control over the model’s output. This document will discuss the concept, its applications, and examples.

Concept of Dynamic Prompting

At its core, dynamic prompting leverages the power of system-level instructions. These are messages with the role “system” that direct the model’s behaviour or set its context. While a traditional approach might have a single system message at the beginning, dynamic prompting can introduce system messages at any point to modify the model’s approach.


  1. Improved Control: Offers better control over the model’s responses based on the evolving conversation.
  2. Context Adaptability: The model can adapt responses to changing contexts or requirements.
  3. Enhanced User Experience: Provides more accurate and contextually relevant answers.


Initial Setup:

Start with a foundational system instruction and a greeting message.

messages = […



Fredric Cliver

Majored in Physics, self-taught and worked in the IT industry as a Dev/Design/Planning for 11 years. And I had run my Startup for 3 years. I fancy a ☔️ 🇬🇧