Limato

Tone Checker

See which register your text actually reads as — then rewrite it into any of seven tones in one click. Free, no signup.

Your text

0 / 750 words

Rewritten text

Paste text and click Check tone to see which register it reads as.

Check tone Free, no signup. Install the extension to switch tone anywhere you type.

How to use the tone checker

Paste up to 750 words and click Check tone. You get back the register your text actually reads as — formal, casual, academic, native, or machine-generated — plus the exact words and constructions that gave it away. Then pick any of the seven tones and the same text comes back rewritten in that register, with the changes highlighted. Your text is processed and returned — never stored.

Detection needs something to read: a three-word fragment has no register. Two or three full sentences is usually enough for a stable verdict.

The seven tones, and what each one actually changes

Each tone is a different rewrite, not a slider on the same one:

• Native — the aggressive one. A measured median 37% shorter (40 drafts, 2026-08-09): textbook openers deleted, hedging cut, the point moved into the first sentence.
• Formal — professional register, no contractions. Changes the vocabulary, keeps the length.
• Casual — everyday words and contractions like it's / we've. What you send a teammate, not a client's legal team.
• Simplify — shorter sentences, aims for B1-level English. For readers who don't share your first language.
• Academic — precise, structured, domain vocabulary. Longer, not fancier.
• Grammar — only fixes grammar, spelling, punctuation; wording untouched. The one to pick when the register is already right. Same engine as the grammar checker.
• Humanize — breaks uniform AI sentence rhythm and compresses 15-35% (measured median 21% shorter across 20 drafts, 2026-08-09). For text that came out of ChatGPT and reads like it — see the AI Humanizer.

What gives a tone away

The verdict comes with evidence: the phrases the register was read from. The usual tells:

• Formal: no contractions, nominalizations ("provide clarification" instead of "clarify"), passive voice, "regarding", "pursuant to".
• Casual: contractions, "stuff", "gonna", exclamation marks, one-word sentences.
• Academic: hedged claims ("suggests that", "tends to"), long noun phrases, citations.
• Machine-generated: sentences all the same length, "delve", "it's worth noting", "in today's fast-paced world", a tidy three-item list in every paragraph.
• Non-native formal: the textbook email kit — "I am writing to", "kindly", "at your earliest convenience".

If what you want is a probability that a machine wrote the text rather than a register, use the AI detector instead — different question, different tool.

Why non-native English lands on the wrong register

Textbooks teach the formal register first, because it is the safe one to teach. The result: most people writing English as a second language have exactly one tone available, and it sits about two notches above anything a native colleague would actually send. The phrases aren't wrong. They're from a business-letter template nobody has used since fax machines.

Before: "I am writing to kindly ask whether it would be possible for you to please be informed about the status of the report at your earliest convenience."

After (Native): "Any update on the report?"

Both are polite. Only one reads like a person wrote it. That kit — "I am writing to", "kindly", "at your earliest convenience", "please be informed", "do the needful" — marks text as non-native far more loudly than a grammar mistake does, and no grammar checker will ever flag it. More of them in 12 ESL patterns native speakers never use.

Which tone fits which channel

• Slack, Teams, internal chat: Casual, or Native if the draft came out stiff.
• Email to a client or a stranger: Native first, then Formal only if the relationship really calls for it. Native is short and direct, which reads as respectful of the reader's time; Formal is long and distant.
• LinkedIn posts and comments: Native. Formal reads as a press release, Casual reads as a stranger being over-familiar.
• Cover letters and legal replies: Formal.
• Docs, READMEs, support replies: Simplify — your reader is probably reading in a second language too.
• Papers, theses, grant applications: Academic.

Tone is not grammar, and not humanizing

Three problems get confused constantly. Grammar is whether the sentence is correct — that's the grammar checker, and it leaves the register alone. Tone is who the sentence sounds like: perfectly correct English can still be too stiff, too chummy, or too academic for the person reading it. Humanizing is a narrow case of tone — text that is correct and well-registered but rhythmically machine-made, which is what the AI Humanizer fixes.

A misplaced comma is a comma problem. "I would like to kindly inform you" is a tone problem. Different checkers, and only one of them is on this page.

Frequently asked questions

More free tools