This repository contains the full system prompts of four major large language models (LLMs):
ChatGPT System Prompt.txtClaude System Prompt.txtPerplexity System Prompt.txtGemini System Prompt.txt
Each file includes the system-level instructions that guide the model’s behavior, capabilities, and interaction design. These prompts offer insight into how each company defines the scope, personality, and control structure of their AI.
| File | Description |
|---|---|
ChatGPT System Prompt.txt |
Prompt used by OpenAI’s ChatGPT, including tool instructions |
Claude System Prompt.txt |
System definition used by Anthropic’s Claude |
Perplexity System Prompt.txt |
Behavior prompt behind Perplexity AI’s conversational agent |
Gemini System Prompt.txt |
Google’s Gemini model behavior configuration prompt |
This repository is created and maintained solely for:
- 📚 Educational purposes – To understand how LLMs are instructed to behave internally.
- 🔬 Non-commercial research – For researchers and developers to compare system prompt design patterns.
- 🧪 Critical analysis and commentary – For prompt engineers and AI practitioners studying model alignment.
All system prompts are shared under the following assumptions:
- They are intended for educational, research, and commentary purposes only.
- This repository is non-commercial and is not affiliated with or endorsed by OpenAI, Anthropic, Perplexity, or Google.
- Contents may include material obtained from publicly visible outputs, documentation, or user interfaces.
This falls under the concept of "fair use", particularly in the context of:
- Education
- Non-commercial research
- Commentary and analysis
If you are the copyright holder of any included content and would like us to update or remove material, please submit an issue or contact the repository maintainers.