Model Guide

ChatGPT

OpenAI's conversational AI - best practices and techniques

Overview

ChatGPT is OpenAI's flagship conversational AI model. It excels at following instructions, engaging in dialogue, and producing human-like text. The latest versions (GPT-4) have strong reasoning capabilities and can use tools.

Available via chat.openai.com and API. GPT-4 offers the best performance but with higher latency and rate limits.

Strengths

Instruction Following

Excellent at following complex, multi-part instructions precisely

Code Generation

Strong coder with good understanding of best practices

Conversation

Natural dialogue and can maintain context well

Tool Use

Can use plugins, browse web, and execute code

ChatGPT-Specific Tips

  • 01
    Use clear structure

    Number your requirements or use bullet points for complex prompts

  • 02
    Leverage system messages

    Set persistent context via system prompts for recurring tasks

  • 03
    Iterate conversationally

    Build on responses with follow-up questions rather than re-prompting

  • 04
    Use delimiters

    Use XML tags or markdown code blocks to separate content

Example Prompt

Effective ChatGPT Prompt
You are a senior software architect reviewing a pull request. Context: - This PR adds user authentication to the existing API - The codebase uses Python/FastAPI - Current tests have 85% coverage Task: 1. Review the code for security issues 2. Check for proper error handling 3. Suggest any performance improvements 4. Verify test coverage is adequate Provide your review in this format: ## Security Issues [Bullet points] ## Suggestions [Bullet points] ## Overall Verdict [Approve/Request Changes]

This prompt works well because it establishes a clear role, provides relevant context, lists specific tasks, and specifies the output format.

← Back to Learn Next: Claude →