Which AI models use system prompts?

All major LLMs support system prompts: ChatGPT (OpenAI), Claude (Anthropic), Gemini (Google), Mistral, LLaMA, and most open-source models. They are also used in API integrations via the system field.

What makes a good system prompt?

Good system prompts are specific about role and task, set clear boundaries on what the model should and should not do, define the output format when relevant, and specify the tone and audience.

AI System Prompt Builder — Generate System

Example Outputs

Code Review

Senior Developer — Review

You are a senior software developer with deep expertise in code quality, security, and best practices. Your primary task is to review and give feedback on code. Respond in a technical and precise tone. Be specific about issues — cite line numbers or patterns when relevant. Prioritise security issues, then correctness, then style.

Customer Support

Support Agent — Friendly

You are a customer support agent representing the company professionally. Your primary task is to answer questions accurately and resolve issues efficiently. Respond in a friendly and approachable tone. Always acknowledge the customer's frustration before solving. Never promise what you cannot deliver.

Content

Writer — Concise

You are a professional writer with expertise in clear, engaging content. Your primary task is to write and edit content. Respond in a concise and direct tone. Prefer short sentences. Avoid jargon unless the audience is technical. Every sentence should earn its place.

Frequently Asked Questions

What is a system prompt?

A system prompt is an instruction given to an AI model before the conversation starts. It defines the model's persona, constraints, tone, and task. Everything in the system prompt shapes how the model responds throughout the session.

Where do I paste this prompt?

In ChatGPT: use the "Custom instructions" feature (Settings → Personalization → Custom instructions) or in the system message field of the API. In Claude: use the system prompt field in the API, or the "System Prompt" box in Claude.ai Projects. In any API call, it goes in the system parameter.

Can I edit the generated prompt?

Yes — the output box is fully editable. The dropdowns give you a solid starting point, but the final prompt is yours to customise. Click anywhere in the output and modify it before copying.

What is the token count for?

Tokens are the units AI models use for pricing and context limits. Roughly 1 token = 0.75 words. System prompts count toward your context window. Most models have a context limit of 4K–200K tokens, so a 100-token system prompt is negligible.

Which AI models support system prompts?

All major LLMs support system prompts via their APIs: OpenAI (GPT-4, GPT-4o), Anthropic (Claude), Google (Gemini), Mistral, Cohere, and all open-source models like LLaMA and Mixtral. They are the standard way to configure AI behaviour in production applications.