AI Reply Generator

What is the AI Reply Generator?

The AI Reply Generator is a free tool that drafts a reply to a message from your notes on what the reply should say. You choose where it will be posted, such as email, Slack, LinkedIn, X or SMS, plus a style and a length. The tool writes one draft that follows that platform's usual conventions.

How to use the AI Reply Generator

1. Paste the message you are answering into Original message. Paste only that message, not the whole thread.
2. Write What the reply should say in note form, such as "Decline politely, offer a call next week".
3. Choose Where it lands, a Style and a Length of Short, Medium or Long. Add Context if the reply depends on something the message does not say.
4. Press Generate reply. The draft appears under Generated Reply. If you are not signed in, press Sign up free to unlock to reveal it.
5. Press Copy text, check the length against the platform's limit, and post it.

How should a reply change from platform to platform?

The same sentence reads differently depending on where it lands. "Thanks for flagging this, I have passed it to the team" is a competent support reply, a stiff Slack message, and a corporate non-answer under a public post. Nothing about the words changed. The venue did.

The Where it lands setting tells the model which platform the reply is for and asks it to match that platform's formatting and norms. That covers whether there is a greeting or a sign-off, how much context can be assumed, and how formal the default register is. An email usually opens with a name. A reply in a Slack thread usually does not, and adding one reads as distance.

What are the character limits for replies on X, LinkedIn and SMS?

A post on X is limited to 280 characters unless the account has a paid subscription. A LinkedIn comment is limited to 1,250 characters. A single SMS holds 160 characters in the standard GSM-7 alphabet, and only 70 once one emoji or non-Latin character switches the message to UCS-2, which is how a short reply turns into several billed messages.

Pick Short under Length for X and SMS, and count the draft before you post it anywhere with a hard limit: the Twitter character counter and LinkedIn character counter do that for those two.

What should you tell an AI reply generator to say?

Tell it the outcome you want, in plain notes. A short reply has room for one idea, and you have to pick it. "Decline, but offer a call next week" produces a different reply than "decline and close it out", and a style setting cannot make that choice for you.

Fragments are fine. What matters is that the outcome is stated plainly, because in a two-sentence reply there is nowhere for an unclear instruction to hide. Facts the sender did not write down, such as "third email this week" or "free tier, asking about SSO", go in the Context field.

For a longer message where several points all have to be answered, the AI response generator takes a list of points and a word count instead.

How do you reply to a public comment or review?

Write it so it still reads well when quoted alone. A reply in a comment thread, a review response or a support ticket can be screenshotted and shared without the message it answered, and a line that is reasonable as a response can look bad on its own.

Before you post, check for four things: sarcasm, an admission of fault you have not confirmed, a mention of another customer, and anything that only makes sense to someone who read the thread above it.

How does the AI reply generator work?

It sends your message, your notes, the platform, the style and the length to an OpenAI model with one instruction: write the reply and nothing else. The draft comes back without an introduction, notes or alternatives around it, so what you copy is what you send.

The notes decide what the reply says, and the other settings decide how it says it. Where it lands asks the model to match that platform's norms, Style sets the register, and Length is passed on as a plain instruction: very concise for Short, balanced for Medium, detailed for Long. Each press of Generate reply writes one draft. If it misses, change the notes rather than the style, because a vague instruction is the usual cause of a reply that says the wrong thing.

The Context field is optional and reaches the model only when you fill it in. Use it for facts the reply depends on that the message does not state: who the sender is, what was promised before, a deadline or a policy. A reply can only say "as we agreed on Tuesday" if Tuesday is in the context, and without it the model fills gaps with general phrasing instead.

Can an AI reply generator write flirty or casual text replies?

It can write casual ones, and you can ask for playful. There is no flirty style in the list, so choose WhatsApp message or SMS under Where it lands, pick Casual or Friendly, and put the tone you want in the notes, such as "light, teasing, ask about Saturday". The notes carry more weight than the style setting, so say exactly what the reply should do.

Edit the draft before you send it. A message to a friend or someone you are dating should sound like you, and a reply that is suddenly more polished than your usual texts reads as odd. Use the draft for the idea, then put it in your own words. The rewording tool gives you other ways to phrase a line you like but would not say.

Which reply style should you choose?

Choose the style for the relationship, then check it against the platform. The eight styles are Professional, Friendly, Empathetic, Direct, Formal, Casual, Assertive and Apologetic, and the same style lands differently in different places: Formal suits an email to a client and reads as sarcasm in a Discord message.

A few pairings work most of the time. Professional or Direct for email and LinkedIn messages. Friendly or Casual for Slack, WhatsApp and SMS. Empathetic for a support ticket from an upset customer. Assertive when you are declining or setting a limit and do not want it read as an opening for negotiation. Apologetic only when the mistake is yours and you mean to say so. To shift the tone of a draft after it is written, paste it into the AI tone changer.

When two styles both seem right, generate the reply in each and compare them side by side. Keep the one you would be comfortable seeing quoted back to you a month from now, which is usually the plainer of the two.