An overhead view of a person writing notes and typing on a laptop at a desk

Photo by RDNE Stock project on Pexels

The gap between a mediocre ChatGPT response and a genuinely useful one is almost always the prompt, not the model. Below is a working set of prompts organized by what you're actually trying to do — writing, work tasks, coding, learning, and everyday life — each one ready to copy, paste, and adapt with your own details. Wherever you see bracketed text like [topic] or [your industry], swap in your specifics before sending.

Writing & content prompts

Work & productivity prompts

Coding & technical prompts

Learning & research prompts

Everyday life prompts

Why these prompts actually work

Every prompt above follows the same underlying pattern covered in our guide to writing better AI prompts: it specifies a concrete output (a length, a format, a tone), gives the model something concrete to react to instead of a blank request, and leaves an obvious slot for your own details. That structure is what separates a prompt that produces something usable on the first try from one that produces generic filler you have to rewrite anyway.

How to adapt these for Claude, Gemini, or any other chatbot

None of these prompts are ChatGPT-specific — the same structure works essentially unchanged on Claude, Gemini, Perplexity, or any other general-purpose AI chatbot, since the underlying skill (being specific about output, tone, and format) is a property of how these models respond to instructions generally, not a platform quirk. If you switch tools, the prompts above are a safe starting point without modification.

Building your own prompt library

The prompts most worth saving are the ones tied to tasks you repeat often — a weekly report summary, a recurring type of email, a standard code review request. Keep a running note (in a plain notes app, or a dedicated AI note-taking tool) of prompts that worked well for you specifically, with placeholders for the parts that change each time. Over a few months this becomes a genuinely useful personal shortcut library, saving you from re-explaining the same context every single time.

Who these prompts are actually best for

These templates give the most value to people who use ChatGPT often enough to have repeat tasks, but not often enough to have already built their own shortcuts. A freelance writer drafting three or four headline sets a week benefits more than someone who opens ChatGPT once a month out of curiosity. Students studying for recurring exams, small business owners writing customer emails daily, and developers who explain the same kind of bug to a chatbot every sprint all fall into this category.

People running highly specialized workflows, like legal document review or medical literature summaries, will find these prompts too generic on their own. They work better as a starting skeleton in those fields, something to modify heavily with domain-specific instructions and constraints rather than use word for word. The same goes for anyone doing creative fiction writing, where a rigid template can flatten a voice that needs more room to breathe.

If you already have a personal system of saved prompts that works for you, this list is less about replacing that and more about filling gaps. Skim the categories, pull the two or three you don't already have an equivalent for, and leave the rest.

Common mistakes people make with these prompts

The most frequent mistake is leaving a placeholder half-filled, typing something like "[your industry, e.g. marketing]" instead of actually replacing it with "marketing." ChatGPT will sometimes try to interpret the bracket literally, and the response comes out confused or generic as a result. Always do a final read-through before sending to make sure every bracket has been replaced with real text.

Another common issue is pasting a prompt meant for one task into a completely different context and expecting the same quality. The coding prompts assume you're pasting actual code or an actual error message. If you skip that step and just describe the problem in vague terms, the response quality drops noticeably because the model has less to work with.

People also tend to under-specify length and tone even when the prompt already has slots for those details. "Write something professional" means something different to almost everyone. Naming a word count, a reading level, or a specific tone (formal, casual, warm, blunt) removes that ambiguity and cuts down on the back-and-forth needed to get a usable draft.

Finally, some users treat the first response as final rather than as a draft to react to. These prompts are strong starting points, not finished products. Reading the output critically and asking for one or two specific changes almost always beats regenerating from scratch.

Limitations of prompt templates

A template prompt can only get you so far because it has no knowledge of your actual constraints unless you type them in. It doesn't know your company's style guide, your professor's grading rubric, or the fact that your manager hates bullet points in emails. Every one of those details has to be added by you, every time, or the output will drift toward generic defaults.

Templates also age. A prompt tuned for a particular model version can behave slightly differently after an update changes how the model interprets instructions, how verbose it defaults to being, or how it handles certain phrasing. None of the prompts here rely on model-specific quirks, but expect some variation in tone and length between models and over time, even with an identical prompt.

There's also a ceiling on what a single prompt can accomplish. Complex, multi-step projects (a full research report, a multi-chapter outline, a codebase refactor) generally need to be broken into a sequence of prompts rather than handled in one shot, no matter how well that one prompt is written. Treat these templates as building blocks for a longer conversation, not a replacement for one.

How to adapt a prompt for your specific situation

Start by identifying which part of the prompt is doing the real work. In most of the templates above, that's the combination of the task verb (write, summarize, compare, quiz) and the constraint that follows it (under 100 words, in a professional report, for a beginner). Keep that structure intact and change only the nouns: the topic, the audience, the format.

If the output still isn't matching what you need, add one constraint at a time rather than rewriting the whole prompt. Adding "avoid technical jargon" or "assume the reader has never used this tool before" to an existing prompt is a smaller, more controllable change than starting over with a brand new one, and it's easier to tell which addition actually fixed the problem.

It also helps to give the model an example of what you want when a description alone isn't landing. Pasting one sentence in the tone or format you're after, then asking it to match that style for the rest, usually works better than trying to describe the style in the abstract. This is especially useful for tone-sensitive writing like bios, cover letters, or customer-facing messages.

ChatGPT pricing at a glance

All of the prompts in this guide work on the free tier, so cost isn't a barrier to using any of them. That said, it helps to know roughly what the paid tiers add if you're using ChatGPT heavily enough to hit free-tier limits. The free plan gives access to a capable model with usage caps that reset periodically and generally covers casual to moderate daily use.

The mid-tier paid plan, aimed at individual power users, typically runs in the range of a modest monthly streaming subscription and removes most of the usage caps, adds access to more capable models, and often includes extras like image generation or longer memory of past conversations. A higher business or team tier usually costs a few times that per seat and adds admin controls, higher usage ceilings, and sometimes priority access during peak load. Enterprise pricing is typically quoted directly to organizations rather than listed publicly, and scales with seat count and support needs.

None of this changes how you'd use the prompts above. The upgrade decision comes down to whether you're regularly hitting usage limits or need a specific paid-only feature, not whether the prompts themselves require it.

Frequently asked questions

Do I need ChatGPT Plus to use these prompts? No — all of these work fine on the free tier of ChatGPT or any comparable chatbot; none require a paid plan or a specific model version.

Why do some of my results still come out generic even with these prompts? Usually because the bracketed placeholders weren't filled in with real specifics — a prompt is only as good as the concrete detail you put into it, so vague substitutions produce vague output.

Can I chain these prompts together in one conversation? Yes, and it often works better than trying to cram everything into one giant prompt — use one prompt to get a first draft, then a follow-up prompt from another category (like an editing prompt) to refine it.

Should I save these somewhere so I don't have to keep coming back? That's exactly the point of the prompt library section above — copy the ones you use often into your own notes so they're one paste away next time.

Do longer, more detailed prompts always produce better results? Not automatically. A prompt padded with irrelevant detail can dilute the instructions that actually matter. The goal is precision, not length. A short prompt with the right constraints usually beats a long one with vague ones.

Will these prompts still work if ChatGPT releases a new model version? In almost all cases, yes. The templates rely on plain instructions about format, tone, and content rather than any model-specific syntax, so they carry over across model updates with little or no adjustment.

What should I do if a prompt gives a good answer but the wrong tone? Keep the same prompt and add a short follow-up specifying the tone you actually wanted, such as "make this warmer" or "less formal, more like a text message." Adjusting an existing draft is faster than starting over with a new prompt.

→ Read the full guide to writing better AI prompts