5 Prompt Formulas That Work Across Every AI Model (Copy & Paste Ready)

Hey there,

Most people treat prompts like they're guessing a password. They throw something at the wall, hope it sticks, and move on.

But prompt engineering isn't magic, it's a science. And once you understand the formulas, you can get consistently better results from every AI model you use.

Today, I'm giving you five battle-tested prompt structures that work across GPT-4, Claude, Gemini, and everything else. Copy them, customize them, and watch your AI outputs improve immediately.

Formula #1: The Context + Task + Format Structure

When to use: Anytime you need a specific output format (emails, social posts, code, reports, etc.)

The Formula:

You are [ROLE/CONTEXT].

Your task is to [SPECIFIC ACTION].

Here's what I need:

• [Requirement 1]

• [Requirement 2]

• [Requirement 3]

Format the output as [DESIRED FORMAT].

Real Example:

You are a senior marketing copywriter specializing in SaaS products.

Your task is to write a compelling email subject line for a product launch.

Here's what I need:

• Subject line should create urgency without being clickbait

• Should mention the key benefit (cost savings)

• Should be under 50 characters

• Should appeal to small business owners

Format the output as: [Subject Line] | [Why it works]

Formula #2: The Chain-of-Thought (Step-by-Step Reasoning)

When to use: Complex problems that require reasoning, analysis, or multi-step solutions.

The Formula:

I need you to solve this problem step-by-step.

Problem: [YOUR PROBLEM]

Please:

1. Break down the problem into smaller parts

2. Analyze each part

3. Explain your reasoning for each step

4. Provide the final solution

Think out loud as you work through this.

Real Example:

I need you to solve this problem step-by-step.

Problem: We're spending $2,000/month on AI subscriptions but only using 3 of the 5 models we're paying for. How should we optimize our AI stack?

Please:

1. Break down what we're paying for vs. what we're using

2. Analyze which models are actually delivering value

3. Explain your reasoning for consolidating or switching

4. Provide a specific action plan with cost projections

Think out loud as you work through this.

Formula #3: The Few-Shot Prompting (Show Examples)

When to use: When you want consistent output style or tone across multiple requests.

The Formula:

I'm going to show you examples of what I want. Then I'll ask you to do the same for new input.

Example 1: Input: [EXAMPLE INPUT] Output: [EXAMPLE OUTPUT]

Example 2: Input: [EXAMPLE INPUT] Output: [EXAMPLE OUTPUT]

Now, do the same for this: Input: [YOUR NEW INPUT]

Real Example:

I'm going to show you examples of how I want customer feedback summarized. Then I'll ask you to do the same for new feedback.

Example 1: Input: "The product is great but the onboarding was confusing. I had to watch three tutorials before I understood how to use it." Output: Positive (product quality) | Negative (onboarding complexity)

Example 2: Input: "Fast, reliable, and the customer support team is amazing. Worth every penny." Output: Positive (speed, reliability, support)

Now, do the same for this: Input: "It's okay but there are cheaper alternatives. The UI feels outdated compared to competitors."

Formula #4: The Role-Playing + Constraint Structure

When to use: When you need the AI to adopt a specific perspective or work within limitations.

The Formula:

You are [SPECIFIC ROLE with expertise].

Your constraints are:

• [Constraint 1]

• [Constraint 2]

• [Constraint 3]

Now, [TASK].

Real Example:

You are a technical writer who specializes in making complex topics simple for non-technical audiences.

Your constraints are:

• Use only words a 10-year-old would understand

• Keep sentences under 15 words

• Use analogies and real-world examples

• Avoid jargon entirely

Now, explain how API rate limiting works.

Formula #5: The Iterative Refinement Loop

When to use: When you're not happy with the first output and need to improve it.

The Formula:

[INITIAL PROMPT]

I like [WHAT WORKED], but I want you to [SPECIFIC CHANGE].

Keep [WHAT TO PRESERVE] and adjust [WHAT TO CHANGE].

Real Example:

Write a LinkedIn post about the benefits of using multiple AI models instead of one.

I like the structure and the data points, but I want you to make it more conversational and less corporate.

Keep the statistics and the practical examples, and adjust the tone to be friendly and relatable instead of formal.

Bonus: The Universal Prompt Checklist

Before you send any prompt, ask yourself:

• Is my role/context clear? (Who should the AI be?)

• Is my task specific? (What exactly do I want?)

• Are my requirements explicit? (What does success look like?)

• Is my format defined? (How should the output look?)

• Have I given examples? (If needed, did I show what I want?)

If you can answer "yes" to most of these, your prompt is solid.

The Prompt Library

Stop reinventing the wheel every time you need to write something.

Geekflare Connect comes with a curated prompt library with thousands of ready-to-use prompts for common business tasks like marketing, sales, customer support, HR, finance, and more. No need to start from scratch.

How it works:

1. Browse the library and find a prompt that matches your task

2. Load it into your chat

3. Customize the variables (change names, dates, topics, etc.)

4. Run it on your preferred AI model

5. Save your own prompts and share them with your team for consistency

📈 Trending in Tech:

Email & SMS marketing that just… works.

Why make things harder than they need to be?

Omnisend gives you powerful email and SMS tools that feel so intuitive, it’s almost boring.

• No learning curve. No drama. Just results.

• $68 ROI for every $1 spent

• Intuitive platform that works with you, not against you

• 24/7 award-winning support

Cheers,

Keval, Editor