Boneyard Tools

AI Emotion Analyzer

See which emotions a piece of text carries. Paste a message, review, or journal entry and the tool detects emotions using a zero-shot text classifier that runs in your browser, scoring joy, love, gratitude, surprise, sadness, fear, anger, disgust and a neutral baseline. Each emotion is scored on its own, so a single text can read as both grateful and surprised. Nothing is uploaded. The model downloads once on first use, then is cached.

How to analyze emotions in text

  1. Paste the message, review, or note you want to read into the box.
  2. Click Analyze emotions and wait a moment while the model loads on first use.
  3. Read the top emotion, then the full ranked list of emotions with confidence scores.

Examples

An excited thank-you message

I just got the job offer and I am thrilled. Thank you so much for believing in me!
Top emotion: joy. Also high: gratitude and love. Low: anger, fear, disgust.

Frequently asked questions

Is my text uploaded anywhere?

No. Nothing is uploaded. The model runs entirely in your browser through WebAssembly, so your text is analyzed on your device and never sent to a server. Only the model itself is downloaded, once, then cached.

How does it detect emotions?

It detects emotions using a zero-shot text classifier that runs in your browser. Zero-shot means the model scores your text against each emotion label (joy, love, anger, sadness, fear, and more) without being specially trained on an emotion dataset, so it stays flexible and runs from one small download.

Which emotions can it detect?

It scores joy, love, gratitude, surprise, sadness, fear, anger, and disgust, plus a neutral baseline. Each emotion is scored independently, so the percentages do not add up to 100 percent and a single text can rank high on more than one emotion at once.

How is this different from sentiment analysis?

Sentiment analysis gives a single positive or negative verdict. Emotion analysis is finer grained: it tells you which specific emotions are present, so you can tell grateful from joyful, or anxious from angry, even when the overall sentiment is the same.

Why does the first run take a moment?

The AI model downloads the first time you use the tool, then is cached by your browser for instant reuse. After that, analysis is nearly instant for short text.

Related tools