Send and Exhale Get the app

Why not just ask ChatGPT to fix my message?

Other AI tools only see the words. Send and Exhale drafts your reply from what really happened.

The short answer

A chatbot makes a message sound nicer. Send and Exhale is built to keep it accurate, brief, and drafted for the people who may read it later, like a judge, guardian ad litem, or mediator.

It's the first question a lot of parents ask, and it's a fair one. If you've ever pasted a message from your ex into a chatbot and asked it to “make this nicer,” you already know it helps with tone. Here's what it doesn't do.

Where general chatbots fall short in co-parenting

  • They take the accusation at face value. Ask a chatbot to reply to “You're always late,” and it may write something like “I'm sorry I've been late.” That's an admission you never made, and it's now in writing.
  • They don't know your history. A chatbot has no idea you were on time the last three pickups, or that you already sent the school calendar on Monday.
  • They over-apologize and over-explain. They're designed to be agreeable. In a high-conflict thread, long and apologetic replies invite another round.
  • They can make things up. AI can fill gaps with confident details that aren't true. In a custody case, one wrong date can cost you credibility.
  • They write for a general audience. Your real audience may be a judge, a guardian ad litem, a custody evaluator, or a parenting coordinator reading months later.
  • You have to know what to ask. Getting a good result from a chatbot takes a careful prompt every time, usually when you're the most upset.

What Send and Exhale does differently

Tuned for court readers

Every draft is built as if a judge or GAL will read it: brief, child-focused, factual, and free of sarcasm and labels.

Claims stay claims

Their accusations are treated as claims, not facts. Your reply is built not to admit something that didn't happen just to sound polite.

Checks your record

Import your OurFamilyWizard or TalkingParents history, and replies can check what the record actually shows.

Shows its work

See What We Read and Why This Reply show you what it used and why. Add more details and it rewrites.

BIFF built in

Brief, Informative, Friendly, Firm, on every reply, without having to remember the method.

Private by design

Your conversations and history are saved on your phone. What's needed for a reply is sent to our AI provider, Anthropic, to draft it. Anthropic doesn't train on it and keeps it only for a limited time.

Same message, two replies

We pasted one hard message into a general AI assistant and asked, “Help me reply to this calmly.” Here is what it wrote, next to a reply built from the parent's record.

What they sent
You were 40 minutes late AGAIN Friday. The kids sat on the curb in the cold. Third time this month, and you never answer my texts. My lawyer is filing for full custody.
A real reply from a general AI assistant, not ChatGPT, given a shorter version of this message (October 2026, results vary)
I hear you, and I'm sorry the kids had to wait on the curb. That's not okay and I don't want that for them. I'm going to make sure pickup is on time going forward. If I'm running behind, I'll text you before the pickup time so you and the kids know. Please let me know if there's anything else that would help.
A reply built from the record
On Friday I arrived at 6:12 for the 6:00 pickup. At 5:52 I texted you that an accident on the highway had me running a few minutes behind. The other pickups this month were on time: Nov 1 at 5:58 and Nov 8 at 6:00. I answered your texts on Nov 3 and Nov 10. If I'm ever running behind, I'll text you before pickup.

Example reply, written from the parent's notes and records. Results vary.

The chatbot reply is polite and kind about the kids. But it answers the feeling and leaves “40 minutes” uncorrected, so it stays on the record. It can't know you texted at 5:52 or arrived at 6:12, because it only knows what you paste in. The reply built from the record answers all three claims with dates and times, without admitting what didn't happen. Send and Exhale also had the parent's notes and records; the chatbot had only the message, which is what most people paste. A chatbot doesn't have your co-parenting records. Send and Exhale starts from your record.

Watch: Why not just ask AI?

Read the transcript

Why not just send your ex's message through AI? Most co-parenting apps are built for receipts, schedules and sending messages. Send and Exhale does one thing: it gives you what to say. Like a friend who sees more: their message, what really happened, and your message history once you add it. Separate on purpose. No link to your messaging app. Nothing goes to your ex by accident. There's no Send button. You change any word, then send it yourself. Just clear, kind replies, drafted from what really happened. Send it calm. Send it accurate. Send it confident. Then exhale. Send and Exhale. Now on iPhone. Free to download. AI-generated person. Example messages. Made-up names. Not legal advice.

How they compare

General AI chatbotSend and Exhale
Built forAnythingReplies to a high-conflict co-parent
Drafted forA general readerA judge, GAL, or mediator reading later
AccusationsMay be accepted or apologized forKept as claims, not admissions
Your historyDoesn't know itCan check your imported records
MethodDepends on your promptBIFF principles every time
Explains choicesRarelyWhy This Reply
Your dataDepends on the service and your settingsYour history is stored on your phone

If you do use a chatbot

  • Delete any apology for something that didn't happen.
  • Check every date, time, and fact against your records.
  • Cut it to a few sentences.
  • Remove anything that sounds like a diagnosis or a label.
  • Don't paste in court documents, your kids' private details, or anything sealed.

Send and Exhale is a writing tool, not legal advice. It doesn't promise any result in your case. You review every draft and decide what to send. ChatGPT is a trademark of OpenAI; Send and Exhale is not affiliated with OpenAI.

Questions parents ask

Can I use ChatGPT to write messages to my ex?

Many parents do, and it can help with tone. Be careful with what it adds. Read every sentence for admissions, apologies for things that didn't happen, and details it made up. Never paste anything you wouldn't want stored by the service you're using.

Is it okay to use AI to write co-parenting messages for court?

Using a writing tool is generally no different from asking a friend to read your message first. What matters is that every word is true and that you stand behind it. You are responsible for what you send. Ask your attorney if you have questions about your case.

What does Send and Exhale do that a chatbot doesn't?

It starts from the message you received, sorts what was said into facts, claims, and your account, checks your imported co-parenting app history, and drafts a brief, friendly, firm reply built for a reader like a judge or GAL. It explains its choices with Why This Reply, and your history stays on your phone.

Want to see the app in action? Watch how it works.

Let Send and Exhale draft the first version

Screenshot the message. Tap Draft My Reply. We’ll draft what to say.

Download on the App Store

Android coming soon.