Tone checker for texts & emails
Not sure how a message will land? Paste it below for a calm, instant read — before you hit send. Built for people who find tone in text hard to gauge, with zero judgment. Nothing you type is sent anywhere; the read runs entirely in your browser.
Want a deeper read — sarcasm, mixed feelings, and a gentler rewrite? That’s the free AI read, available with an account (3 a day).
Open the AI read →How the same words can land
Small signals do a lot of quiet work. Here’s how a few common messages are often received — and a warmer way to say the same thing.
“Per my last email, this is due today.”
Can read as: Often received as pointed or exasperated — "per my last email" is one of the most reliably passive-aggressive-sounding phrases in work writing, even when you only meant to reference context.
Warmer: “Just resurfacing this — it’s due today, so flagging in case it slipped. Thanks!” — Referencing that you already said it puts the reader on the back foot; naming the thing plainly doesn’t.
“Fine.”
Can read as: A lone word ending in a full stop reads as serious or annoyed to a lot of people — the punctuation does the emotional talking, even if you feel neutral.
Warmer: “Fine by me! 🙂” — Dropping the full stop and adding a word removes the "something’s wrong" signal.
“You need to send me the file ASAP.”
Can read as: Clear, but the stacked urgency ("need to" + "ASAP") can read as demanding, especially without a greeting or reason.
Warmer: “Could you send the file when you get a moment today? It unblocks the review — thanks!” — A reason plus a soft deadline keeps it direct without the edge.
“k”
Can read as: To some people this is just efficient. To others "k" is the coldest reply in the language. Context and your history decide — but the risk of a chill read is real.
Warmer: “sounds good!” — Same meaning, none of the "I’m done talking" undertone.
“I just wanted to maybe suggest, if that’s okay, that we possibly reconsider?”
Can read as: The pile-up of hedges ("just", "maybe", "if that’s okay", "possibly") can undersell a fair point and read as unsure.
Warmer: “I’d suggest we reconsider this — here’s why.” — Stating it plainly usually reads as more confident, not less kind.
Practice writing a reply
Reading tone is half the skill; writing it is the other half. Draft a reply to each situation, then reveal a strong example and see the difference a few words make. In the app, you can also get gentle AI feedback on your own wording.
In a team channel, a coworker summarized an idea as theirs — but it was one you raised last week. You want to set the record straight.
Draft a reply that corrects the record without sounding combative or petty.
Show a strong example
Glad this is landing well! Quick note for the thread — this builds on the approach I floated last week, so happy to pair with anyone picking it up. Excited to see it move.
Why it works
- Opens warm, so it doesn’t read as an attack.
- States the fact plainly ("I floated last week") without accusing.
- Offers to help, which signals collaboration over territory.
Easy to overdo
- Over-apologizing ("sorry to be that person…") hands away the point you’re making.
- Going fully indirect ("interesting, where did this come from?") can read as passive-aggressive.
You emailed a colleague three days ago and haven’t heard back. You need an answer to move forward, but you don’t want to sound impatient.
Draft a follow-up that nudges without the passive-aggressive edge.
Show a strong example
Hi Sam — floating this back to the top of your inbox in case it got buried. No urgency today, but I’ll need a yes/no by Thursday to keep the launch on track. Thanks!
Why it works
- Names the real reason for the nudge (a deadline), so it’s not just pressure.
- Avoids "per my last email" and "just circling back" — the phrases most often read as pointed.
- Gives a concrete date, which is kinder than a vague "ASAP".
Easy to overdo
- "Not sure if you saw my email…" almost always reads as a dig, even when sincere.
- A one-line "?" reply can feel curt enough to sting.
You’ve been invited to a recurring meeting that isn’t a good use of your time. You want out without seeming uncooperative.
Draft a decline that’s clear but leaves the door open.
Show a strong example
Thanks for including me! I don’t think I’m adding much in this one, so I’m going to step back to protect focus time — but please tag me on the notes or grab me if a decision needs my input.
Why it works
- Says the actual decision ("I’m going to step back") instead of hinting at it.
- Gives a reason that isn’t about the other person.
- Keeps a clear on-ramp back in, so it doesn’t read as a door slam.
Easy to overdo
- Endless hedging ("if that’s okay, no worries either way, sorry…") can read as more awkward than a clean no.
- No reason at all can feel abrupt in a work context.
A manager messages you at 9pm asking for something "when you get a sec". It can wait until morning, and you’d like to keep your evenings.
Draft a reply that holds the boundary warmly.
Show a strong example
On it first thing tomorrow — I’m offline for the evening but this is top of my list in the morning. Shout if it’s actually urgent tonight and I’ll jump on it.
Why it works
- Confirms you’ll handle it, so the boundary doesn’t read as brushing them off.
- States the boundary as a fact, not an apology.
- Leaves a genuine escape hatch for real emergencies.
Easy to overdo
- Silence until morning can read as ignoring them; a one-liner now prevents that.
- Over-explaining why you’re offline can sound defensive.
A close friend texts just "hey can we talk later?" with no other context. Your brain immediately assumes the worst.
Draft a reply that checks in without spiralling or sounding defensive.
Show a strong example
Sure — I’m free after 6. All good on my end; hope everything’s okay with you?
Why it works
- Answers the actual question (a time) instead of interrogating the vibe.
- Names your own calm ("all good on my end") without demanding reassurance.
- Leaves space for them to share — or not.
Easy to overdo
- "Why? Did I do something?" reads your anxiety onto them and can create tension that wasn’t there.
- A message with no context like this is genuinely unreadable — it’s usually neutral, and asking beats guessing.
A friend invites you to a big party. You’re not up for it, but you don’t want them to think you’re blowing them off.
Draft a decline that’s honest and keeps the friendship warm.
Show a strong example
Ah I’m going to sit this one out — not up for a big crowd this week. Really glad you asked though, and I’d love to do something lower-key soon. Have the best time!
Why it works
- Declines clearly so there’s no mixed signal to decode.
- Names a reason that isn’t about them.
- Offers a real alternative, which reads as "I want you, not the event".
Easy to overdo
- "Maybe, we’ll see…" leaves them managing your uncertainty and can read as a soft brush-off.
- Three apologies can make a simple no feel like a bigger deal than it is.
You snapped at a friend yesterday when you were stressed. You feel bad and want to repair it.
Draft a message that owns it without collapsing into self-criticism.
Show a strong example
Hey — I was short with you yesterday and that wasn’t fair to you. I was stressed but that’s on me, not you. Sorry. Still on for Saturday if you are?
Why it works
- Names the specific thing, so it’s a real repair, not a vague "sorry".
- One clean apology — not a spiral.
- Moves back toward normal ("still on for Saturday?"), which signals the friendship’s fine.
Easy to overdo
- Piling on self-blame ("I’m the worst, you deserve better…") makes them comfort you instead of the other way round.
- Explaining the stress at length can read as excusing rather than owning.
You sent a long, excited text about your weekend and got back a flat "nice". You’re not sure if they’re annoyed or just busy.
Draft a reply that doesn’t over-read the "nice".
Show a strong example
Haha anyway — how’ve you been? Feels like it’s been a minute.
Why it works
- Doesn’t treat one short reply as a verdict — a single "nice" is famously unreliable.
- Gently opens the door back up instead of demanding to know what’s wrong.
- Keeps your own tone light, which doesn’t force a mood onto them.
Easy to overdo
- "You seem off, did I annoy you?" can manufacture a problem from a busy-thumb reply.
- Matching the flatness back ("k") can start a cold spiral neither of you meant.
Questions people ask
How can you tell the tone of a text message?
You can’t hear tone in text, so readers infer it from small signals: length (a one-word reply can read as cold), punctuation (a lone "Fine." often reads as annoyed), word choice (hedges, sarcasm, pointed phrases like "per my last email"), and capitalization. This tool flags those signals so you can see how a message might land — but tone is genuinely ambiguous in text, so treat every read as a probability, not a fact.
Why do my texts sound rude or blunt when I don’t mean them to?
Usually it’s a mismatch between how you feel writing it and how it reads without your voice, face, or timing attached. Short replies, full stops after single words, and skipping softeners ("thanks", "when you get a chance") can all read colder than intended. None of that means you did anything wrong — it just means the medium strips out the warmth you’d carry in person, and a few words can add it back.
Is this useful for autistic adults or people with social anxiety?
Many autistic adults and people with social anxiety find text tone hard to gauge — either reading threat into neutral messages, or worrying their own messages come across wrong. This tool is built for exactly that: it makes the invisible signals explicit and calm, and it’s deliberately careful not to over-claim. It’s general practice, not a clinical or diagnostic tool, and not a substitute for professional support.
What’s the difference between the casual and professional settings?
The calibration changes what counts as notable. In casual mode, a one-word reply or a trailing "…" is weighed for closeness and warmth. In professional mode, the same tool watches for things like missing greetings, stacked urgency, over-hedging, and passive-aggressive email phrases — the patterns that land differently at work than in a group chat.
Does my message get sent anywhere or stored?
The instant read runs entirely in your browser — nothing you type is sent to a server. If you choose the optional AI deep-read (a signed-in feature), that specific message is sent to the AI model to analyze and is not stored afterwards. The choice is always yours, and the free on-device read never leaves your device.
Can a tone checker be wrong?
Yes — often, and it’s important to know that. A tool can only see words on a screen. It can’t know your history with the person, your actual voice, in-jokes, or the wider context. It’s a second pair of eyes to think alongside, not a verdict. When it really matters, the most reliable move is still to ask the person directly rather than guess.
Reading tone is a skill — and skills grow with practice.
Try short, friendly scenarios on tone,
subtext, and more — free to start, no sign-up.
Related: the cue glossary · what a soft no looks like · the blog.