Quick Answer
A system prompt is the standing instruction that sets an AI's role, rules, tone, and output format before you ask it anything. It's the assistant's job description, applied to every message in a conversation. Write it clearly, keep it focused on one job, and the AI behaves far more predictably.
To write a system prompt, you set the AI's role, rules, tone, and output format up front, before you send any actual task, so the model follows those instructions on every message (Anthropic; arXiv). That's the whole idea. A system prompt is the assistant's job description. The user prompt is the individual task you hand it within that job. Get the system prompt right and you stop repeating yourself, and the AI stays consistent across a long conversation. This guide covers what a system prompt is, how it differs from a user prompt, how to structure one, and where Claude and ChatGPT treat them differently. If you're new to prompting, start with our prompt engineering basics.
What Is a System Prompt?
A system prompt is the stable instruction layer that shapes how an AI responds throughout a conversation. According to PromptLayer, the system prompt specifies the model's role, constraints, output format, and communication style. Think of it as the AI's permanent job description: who it is, how it should talk, what it must always do, and what it must never do.
You've seen this in action even if you've never written one. When a custom GPT or a company chatbot stays perfectly on-brand no matter what you ask, that's a system prompt doing its job in the background. It sets the frame once, and every reply inside that conversation inherits it.
What's the Difference Between a System and User Prompt?
This is the distinction that trips people up, and it's worth nailing. The system prompt is the fixed layer. The user prompt is the live request. PromptLayer puts it simply: rules you need the model to follow regardless of what the user says go at the system level, and tasks that change with each request go at the user level.
| System prompt | User prompt |
|---|---|
| Fixed across the conversation | Changes every turn |
| Role, tone, rules, format, limits | The specific task or question |
| "You are a support agent for Acme" | "Reply to this refund request" |
So the system prompt is the AI's permanent job description, and the user prompt is an individual task within that job. Both matter. But the system prompt is where you set the behaviour you don't want to keep repeating.
Advertisement
Why Does the System Prompt Matter So Much?
Because it controls everything downstream. A good system prompt means you set your rules once and they hold across dozens of messages, instead of adding "keep it under 100 words and stay formal" to every single request. It's the difference between an assistant that drifts and one that stays on task.
There's a cost angle too, and it's easy to miss. In API workflows, the system prompt gets sent along with every call, so it counts toward token usage and cost on each one. As the system prompt guides point out, a 2,000-token system prompt across 10,000 calls creates real spend. So a tight, focused system prompt isn't just clearer. It's cheaper.
How Do You Write a Good System Prompt?
The best system prompts are structured, not rambling. According to PromptArch's 2026 best practices, a well-structured system prompt contains a handful of clear blocks. Here's the shape to aim for:
- Identity and persona. Who the assistant is and who it's talking to. Keep it tight, two or three sentences.
- Task scope. What it's here to do, and just as important, what it's not here to do.
- Rules and constraints. The hard limits: tone, word count, what to avoid, any facts it must not invent.
- Tool-use rules. If it can call tools or search, when it should and shouldn't.
- Refusal behaviour. How to handle requests it can't or shouldn't fulfil.
- Output format. The exact shape you want back, whether that's JSON, a table, or three short paragraphs.
The golden rule that ties it all together: each prompt should do one thing well. PromptArch is blunt about this. If your system prompt is trying to make the AI a customer service rep, a financial analyst, and a creative writer at the same time, the quality of all three suffers. Pick one job. And don't leave anything open to interpretation. Tell the AI exactly what to deliver and exactly what to avoid. For more on clear instructions, our guide on how to write better prompts goes deeper.
How Do System Prompts Differ Across Claude and ChatGPT?
The same system prompt doesn't behave identically on every model, and knowing the difference saves you headaches. PromptArch's testing highlights two big ones.
Claude has a dedicated system parameter that's genuinely privileged, meaning instructions there are harder to override from the user turn. Claude also follows XML-like tags very reliably and rewards structured prompts, so wrapping sections in tags like <rules> and <format> works well.
ChatGPT (GPT-5 and GPT-4o) has a system role too, but it's less privileged than Claude's, and a sufficiently determined user prompt can override it. For hard contracts you want enforced no matter what, PromptArch suggests using developer messages and response-format constraints rather than relying on the system role alone. In plain terms: on Claude, the system prompt is close to law. On ChatGPT, treat it as strong guidance and back up your non-negotiables with format constraints.
What Are the Common System Prompt Mistakes?
Most weak system prompts fail in the same few ways. Watch for these:
- Cramming in too many jobs. One role, one focus. A prompt that wants the AI to be five things does none of them well.
- Being vague. "Be helpful and professional" tells the model almost nothing. Spell out the tone, the length, and the format.
- Skipping the output format. If you don't say what shape you want back, you'll get a different shape every time.
- Forgetting refusal rules. Say what the AI should do when it can't help, or it'll improvise, sometimes badly.
- Letting it bloat. Every token you add costs money on every call and can dilute the important instructions. Cut anything that doesn't change behaviour.
- Assuming it transfers between models. Test your prompt on the model you'll actually use, since Claude and ChatGPT weight the system layer differently.
Fix those and your system prompts get shorter, cheaper, and far more reliable. Write it once, test it, trim it, and let it do the heavy lifting so your user prompts can stay simple.
What Else Do People Ask?
What is a system prompt in AI?
A system prompt is the standing instruction that tells an AI its role, rules, tone, and output format before you give it any task. It's the assistant's job description, and it applies to every message in that conversation. The user prompt is the individual request you send within that job.
What is the difference between a system prompt and a user prompt?
The system prompt sets the model's role, constraints, output format, and style, and it stays fixed across the conversation. The user prompt is the per-request input, the specific task or question for that single turn. Rules the model must always follow go at the system level; tasks that change each time go at the user level.
How long should a system prompt be?
Long enough to be clear, short enough to stay focused. Keep the identity to two or three sentences, and give it one job to do well. If a system prompt tries to make the AI a support rep, an analyst, and a creative writer at once, the quality of all three drops. Cut anything that doesn't change how the model behaves.
Do system prompts count as tokens?
Yes. In API workflows, the system prompt is sent along with every call, so it counts toward token usage and cost each time. A 2,000-token system prompt across 10,000 calls adds up to real spend, so keep it tight and cut anything that isn't earning its place.
Can a user prompt override a system prompt?
It depends on the model. Claude's system parameter is genuinely privileged, so instructions there are hard to override from the user turn. In GPT-5 and GPT-4o the system role is less privileged and a determined user prompt can override it, so hard rules are better enforced with developer messages and response-format constraints.
Sources: PromptLayer on system vs user prompts; PromptArch's 2026 system prompt best practices for structure, model differences, and token cost. Both linked above.