Formal, neutral or casual tone detection

Writers and communicators can review a message before sharing it when they need to confirm how formal or casual its language feels.

Identify whether a text sounds formal, neutral or casual, making its current register clear before you adjust the wording for your audience.

How to use

Paste the message you want to examine into the input field, or load a text file when you are working from an existing draft.
Run the analysis to place the text between formal, neutral and casual language.
Read the result with the intended audience in mind, since the same wording may feel neutral to a colleague and too casual for a client.
Adjust your chosen phrases, then run the analysis again if you want to check the new balance.

Example

Take an email that opens with a relaxed greeting and continues with a very direct request. The analysis helps show whether the whole message leans casual or neutral, and you decide whether that register fits the recipient.

Input

Hi, can you send me the file right away? Thanks!

Limits and formats

Perceived tone depends on context, relationship and channel, none of which the tool knows.
Detection works from language signals, not from the writer's actual intention.
The file picker does not read images, PDF files or presentation documents.
A very short message may give a less stable indication than a longer piece of writing.

Where processing happens

Tone evaluation happens in the browser. Your message stays on your device and is not passed to another party, so you can test a sensitive reply before sending it.

Frequently asked questions

How is the tone decided?

The tool looks at language signals to suggest a position between formal, neutral and casual. It cannot read minds or know your relationship with the reader, so the result is only an indication.

Why does my text seem misclassified?

A message that mixes registers can be hard to place. Greetings, vocabulary and sentence length all influence the outcome, so review the full context before deciding whether the label fits.

Can I analyse an imported file?

Yes, you can load a compatible text-based or structured file such as CSV, JSON, MD, XML, YAML or YML. Images and PDF documents are not supported by the picker.