Skip to content

Free tool · Prompts

AI prompt generator

Describe what you want and get a ready-to-paste prompt with a role, context, steps and format: 1 free run a day on GPT-6 Luna, no account needed. Signed in, run the prompt on any of 15 models right away.

1 free run a day, no account needed

GPT-6 Luna

Up to 1,200 characters. No files, tools or web search in a free run. One free run a browser, and one a network each day, shared with the free message on our home page.

Try an example:

By running it, you agree to the Terms and Privacy Policy and confirm you’re 18 or older. We keep what you send and its answer for 7 days, or in your account if you sign up.

Test runs checked . Updated .

AI prompt generator: three examples

Three inputs and the prompt a free run gave each, made once on September 28, 2026 with GPT-6 Luna and shown as it came back. The example buttons in the box above show the same.

A trip plan

Plan a 5-day trip to Japan for two people in spring, mid-range budget, we like food and walking.

Prompt:
Act as an experienced Japan travel planner. Plan a practical 5-day spring trip in Japan for two travelers who enjoy food and walking and have a mid-range budget. Use [travel dates] and [departure city] if provided; if not, make reasonable assumptions and clearly state them. Choose a route that avoids spending too much time in transit.

1. Suggest the best cities or neighborhoods for this trip and briefly explain why.
2. Create a day-by-day itinerary with walkable activities, food experiences, and realistic travel times.
3. Recommend mid-range lodging areas and ways to keep costs manageable.
4. Give an estimated total budget for two, broken down by lodging, food, local transport, intercity travel, and activities. State the currency and key assumptions.
5. Add practical spring travel tips and note anything that should be checked or booked in advance.

Format the answer with a short overview, a 5-row itinerary table, a budget breakdown, and practical tips. Be specific but concise, and don’t invent current prices or availability; flag estimates that need verification.

Fill in: [travel dates], [departure city]
Meeting notes to an email

Turn my messy meeting notes into a follow-up email with decisions and action items.

Prompt:
<context>
You are an expert meeting-note editor. Turn the notes below into a clear, professional follow-up email for [recipient or audience]. The email should help attendees remember what was decided and what happens next.
</context>

<task>
1. Review the notes and identify confirmed decisions and action items.
2. Draft a concise, friendly email that preserves the notes’ meaning. Don’t invent details or present suggestions as decisions.
3. For each action item, include the owner and deadline only if the notes specify them. Mark missing information as “Not specified.”
4. If something is unclear or contradictory, flag it briefly rather than guessing.
</task>

<meeting_notes>
[Paste your messy meeting notes here]
</meeting_notes>

<output_format>
Subject: [Concise subject line]

Hi [greeting],

[Brief opening and meeting recap.]

Decisions
- [Decision, or “None recorded”]

Action items
- [Action — owner: … — deadline: …]

[Brief closing]

Best,
[Your name]

Clarifications needed: [List any unclear or contradictory points, or “None.”]
</output_format>

Fill in: [recipient or audience], [Paste your messy meeting notes here], [Your name]
Explain a coding error

Explain a Python error message to me like I'm a beginner and tell me how to fix it.

Prompt: You are a patient Python tutor helping a beginner understand and fix an error. Explain things in plain language, avoid unnecessary jargon, and define any technical terms you use.

I’ll provide an error message and, if available, the relevant code and what I expected to happen. Follow these steps:

1. Identify the important part of the error message and explain what it means.
2. Describe the most likely cause in the provided code. If there isn’t enough information to be sure, say what’s uncertain rather than guessing.
3. Give clear, numbered steps to fix the problem.
4. Show corrected code when the supplied code is sufficient. Explain the key change.
5. If you need more information to diagnose it, specify exactly what I should share.

Use this output format:

What the error means: [beginner-friendly explanation]
Likely cause: [cause or what’s uncertain]
How to fix it:
1. [step]
2. [step]
Corrected code: [code, or say what additional information is needed]
Why this works: [brief explanation]

Example of the style:
What the error means: Python tried to turn some text into a whole number, but the text wasn’t a valid number.
Likely cause: The value contains letters instead of digits.
How to fix it:
1. Check what value is being passed to `int()`.
2. Make sure it contains a number, or handle invalid input.
Corrected code: `age = int("12")`
Why this works: `"12"` contains digits that Python can convert to an integer.

Here’s my error and code:
Error message: [paste the full error message]
Relevant code: [paste the relevant code]
What I expected to happen: [describe what you expected, if known]

Fill in: [paste the full error message], [paste the relevant code], [describe what you expected, if known]

Which AI model is best at this? We tested every one

Every one of the 15 models we ran passed all 5 of our AI prompt generator test inputs, so choose on speed and price: GLM 5.3 was quickest, at 1.8 s a run on average. GPT-6 Luna, the model a free run is on, took 3.4 s.

Run on September 28, 2026: 5 inputs on 15 models, each with this tool's own instructions.

Each model's results on the tool's test inputs
ModelPassedHard onesMessages used on ProCost per runTime per run
GPT-6 LunaOpenAIThe free run's model5 of 52 of 21 each, of 60 a day on Pro$0.00013.4 s
GLM 5.3 FlashZ.ai5 of 52 of 21 each, of 60 a day on Pro$0.00022.7 s
DeepSeek V4.1 FlashDeepSeek5 of 52 of 21 each, of 60 a day on Pro$0.00042.4 s
Gemini 3.8 FlashGoogle5 of 52 of 21 each, of 250 a month on Pro$0.00062.6 s
GLM 5.3Z.ai5 of 52 of 21 each, of 250 a month on Pro$0.00151.8 s
Claude Haiku 4.5Anthropic5 of 52 of 21 each, of 250 a month on Pro$0.00174.1 s
DeepSeek V4 ProDeepSeek5 of 52 of 21 each, of 250 a month on Pro$0.004813.5 s
Grok 4.7xAI5 of 52 of 21 each, of 250 a month on Pro$0.010823.1 s
GPT-6 SolOpenAI5 of 52 of 21 each, of 125 a month on Pro$0.00275.0 s
Kimi K3Moonshot5 of 52 of 21 each, of 125 a month on Pro$0.003623.7 s
Claude Sonnet 5Anthropic5 of 52 of 21 each, of 125 a month on Pro$0.00586.3 s
Gemini 3.1 Pro (preview)Google5 of 52 of 21 each, of 125 a month on Pro$0.007210.8 s
Claude Opus 5.5Anthropic5 of 52 of 21 each, of 62 a month on Pro$0.02019.5 s
GPT-6 AstraOpenAI5 of 52 of 21 each, of 31 a month on Pro$0.00935.3 s
Claude Fable 5.1Anthropic5 of 52 of 21 each, of 31 a month on Pro$0.02947.6 s
Passed: of the 5 test inputs, how many answers passed their check (an answer the provider failed to give isn't counted). Messages used: what each run counted as on Pro, and Pro's count on the model. Cost: what OpenRouter charged us per run, on average; in llmwise you pay per message, not per token. Time: from sending to the whole answer.

Side by side: meeting notes to action items, in detail

The same input on GPT-6 Luna and GLM 5.3 Flash, as each answered it, unedited.

Write a ready-to-paste AI prompt for this. Give it a role, the context, the steps and the output format, with anything only I know in [brackets]. Make it detailed: up to 300 words, with an example of the output.

---
Turn my messy meeting notes into a table of action items, each with an owner and a due date, and flag anything with no owner.
GPT-6 Luna
Prompt:
You are an expert meeting-notes analyst. I will provide messy notes from a meeting. Turn them into a clear, accurate list of action items for the people responsible for following up.

Steps:
1. Read the notes and identify each distinct task, commitment, or follow-up.
2. For each item, extract the owner and due date only when the notes explicitly provide them. Do not guess or invent missing details.
3. If an item has no clearly stated owner, mark its owner as “Unassigned” and flag it. If the due date is missing or unclear, write “Not specified.”
4. Keep action items concise and preserve important context. Do not turn general discussion into tasks unless the notes indicate a follow-up.
5. If a detail is ambiguous, briefly note the ambiguity in the final column rather than resolving it yourself.

Output only a Markdown table with these columns: Action item | Owner | Due date | Flag or note. Use “No owner” in the Flag or note column for every unassigned item.

Example format:
| Action item | Owner | Due date | Flag or note |
| [Brief task from the notes] | [Person named in the notes] | [Date stated in the notes] | — |
| [Brief task from the notes] | Unassigned | Not specified | No owner |

Meeting notes:
[paste your messy meeting notes here]

Fill in: [paste your messy meeting notes here]
GLM 5.3 Flash
Prompt: You are an experienced project coordinator who turns rough meeting notes into clear action plans. I will paste messy meeting notes from [meeting name/date]. Your job is to extract every action item and organize them into a table.

Steps:
1. Read the notes carefully and identify every task, decision requiring follow-up, or promise someone made.
2. For each action item, determine the owner (the person who will do it) and the due date, based only on what the notes say. Do not invent names or dates.
3. If a task has no stated owner, put "UNASSIGNED" in the owner column and add a flag like "⚠ needs owner" in a Notes column. Do the same for missing due dates ("⚠ needs date").
4. If the same task appears multiple times, merge it into one row and note it was repeated.
5. After the table, list a short "Follow-ups" section: items needing an owner or date, and any ambiguous tasks where you weren't sure what was meant, quoting the original wording.

Output format: A markdown table with columns: # | Action Item | Owner | Due Date | Notes. Keep each action item to one clear sentence starting with a verb. Then the "Follow-ups" bullet list. Aim for completeness over brevity; include every task you find.

Example output:
| # | Action Item | Owner | Due Date | Notes |
|---|---|---|---|---|
| 1 | Draft the Q3 budget summary | Priya | Mar 14 | |
| 2 | Send updated contract to vendor | UNASSIGNED | Mar 20 | ⚠ needs owner |

Follow-ups:
- Task 2 has no owner in the notes ("someone should send the contract?").

Here are my notes: [paste your meeting notes here]

Fill in: [meeting name/date], [paste your meeting notes here]

How the inputs are scored

  • Key facts: the answer must state each fact the input turns on (a number, a name, a date or a condition), as the input gives it; how it's worded isn't judged.
  • Rubric (graded): the grader model scores it from 1 to 5 on each published criterion; it passes with an average of 4 or more and no criterion under 3.

The inputs: a family meal plan on a budget, interview practice, for chatgpt, sorting customer reviews, for claude, meeting notes to action items, in detail, and a birthday toast with nothing invented. Every model gets the tool's own instructions and up to Pro's reply limit. The method is the same as our other test runs.

How to use the AI prompt generator

  1. Say what you want the AI to do, in a sentence or two. Up to 1,200 characters.

  2. Pick the model it's for and how detailed the prompt should be.

  3. Press “Write my prompt”. Without an account you see its start (1 free run a day, no account needed).

  4. Fill in the [brackets], then paste it into any chat. Signed in, run it right away on any model.

Tips for a better answer

  • Say who it's for

    “A cover note for a sales job at a small bike shop” gets a better prompt than “a cover note”. The more the generator knows, the fewer brackets you fill.

  • Ask for the format you need

    If you need a table, a list or a word count, say so in your request; the generated prompt will pin it down.

  • Iterate in the chat

    Run the prompt, and if the answer misses, tell the model what's wrong in the same chat instead of starting over: it keeps the context.

  • Save your best prompts

    Keep the prompts that work in a note. The next time you need one, change the brackets, not the whole thing.

What a good prompt looks like

  • A role and an audience

    A good prompt says who the model should be, what it's for and who the answer is for. “You're a travel planner writing for a couple on a mid-range budget” beats “plan a trip”.

  • The format, stated

    It spells out what the answer should look like: a table, five bullet points, 200 words, a day-by-day plan. The format is what makes an answer usable without a second message.

  • Blanks for your details

    Everything only you know (your budget, your dates, your code) is left as a [blank] for you to fill in, not guessed.

  • No padding

    It's short enough to read. A prompt that runs to a page usually repeats itself; the model does better with the few instructions that matter.

What's free, and what signing up adds

  • Without an account: 1 free run a day, on GPT-6 Luna, shared with the free message on our home page. It takes up to 1,200 characters, writes up to about 600 tokens (roughly 450 words), and shows you the start of the prompt.

  • Signing up is free, with an email link and no card. The prompt you ran comes with you as your first chat, and you get 5 free messages on almost every model (all but Claude Fable 5.1, Claude Opus 5.5, and GPT-6 Astra).

  • Signed in, the tool runs on the model you pick, on your plan, shows the whole answer, and you can carry on in the chat with any model, or ask another model the same thing.

More free tools are on the tools page, and what each plan gives is on pricing.

Questions

Is the prompt generator free?

Yes. Without an account you get 1 run a day on GPT-6 Luna, shared with the free message on our home page: you see the start of the prompt, and the rest when you sign up. Signing up is free and adds 5 free messages on almost every model in llmwise (all but Claude Fable 5.1, Claude Opus 5.5, and GPT-6 Astra), with no card.

Does it work for ChatGPT, Claude and Gemini?

Yes. Pick ChatGPT, Claude or Gemini and the prompt is written the way that model reads best; “Any model” works everywhere. In llmwise you can run the same prompt on all 15 models and compare.

What are the parts in brackets?

Words in [square brackets] are the details only you know: your dates, your budget, your text. Replace each one before you paste the prompt, or the model will guess.

Can it improve a prompt I already wrote?

Paste your own prompt into the box with a line such as “Improve this prompt”. It keeps what you asked for and adds the role, context and format it was missing.

Where can I find ready-made prompts?

Our prompt library has tested prompts for common jobs, each with what it costs to run on every model in llmwise.

Which AI is best for this, in your tests?

Every one of the 15 models we ran passed all 5 of our AI prompt generator test inputs, so choose on speed and price: GLM 5.3 was quickest, at 1.8 s a run on average. GPT-6 Luna, the model a free run is on, took 3.4 s. The table on this page has every model's result, what a run counts as on Pro, and the time it took.

Claude, GPT, Gemini, DeepSeek, Grok, Kimi, and GLM, in one chat.

See what a message costs before you send it. Free is 5 messages to try; sign in with an email link, no password or card.