What patterns are hiding in your conversations?

Spend enough time in a group chat and you start to sense things. Your messages seem to disappear more often than others'. Someone else gets a reply every single time. Chat Analyzer finds the evidence for those hunches in your actual message history.

Upload an exported chat from KakaoTalk, LINE, WhatsApp, WeChat or Telegram to review each participant’s observable communication patterns and the direction of interactions. Results are reference information, not a diagnosis of personality or intent.

• Ignore rate & brush-off rate — how often you speak and nobody responds, and how often you only get a laugh back
• Communication patterns — questions, topic shifts, initiative and response balance based on observable messages
• Group dynamics — who engages with whom, and who is quietly left on the edge
• Chat search — ask "what was that place someone mentioned?" and AI digs it up

Why does the same group chat feel so different to each person?

A group chat looks like one shared room, but the experience inside it is rarely the same for everyone. One person drops a single line and three replies appear. Another writes a careful paragraph and the conversation moves on as if nothing was said. We usually explain this away as personality, but the chat log itself holds a far more specific answer.

Read the transcript closely and patterns emerge. Certain people reliably draw responses; others do not. Some receive replies whenever they speak but rarely answer anyone else. That directionality is something participants sense but almost never say out loud.

Chat Analyzer surfaces that structure with numbers attached. Nothing here is a mood judgement. Every figure comes from counting the actual order of messages and the responses that followed them.

What the analysis covers

Ignore rate measures how often you spoke and the topic moved on without anyone picking it up. Brush-off rate tracks replies that were only laughter or a short answer. These are conversation signals, not proof of another person’s feelings.

Participant summaries organize question frequency, initiative, response style and message length from the supplied conversation. Each interpretation is accompanied by the message pattern that supports it.

You can also review peak activity hours, characteristic vocabulary, topic changes and interaction directions between participants.

How to read your results

Treat the output as a mirror rather than a diagnosis. A high ignore rate does not mean something is wrong with you. Room dynamics, how close the participants are, and the period the messages cover will all move the same person is numbers substantially.

The most useful approach is to focus on what you can actually change. If your messages get buried, try breaking them into shorter lines, or raising a topic when the room is already active. Those two adjustments alone often shift response rates noticeably.

And if your results come back flattering, take that at face value too. Finding a flaw in everyone would not make the analysis more accurate.

How your data is handled

Uploaded chat files are never stored on our servers. Message counts, ignore rates, hourly distribution and every other statistic are computed inside your own browser, so the file itself never leaves your device.

When AI commentary is generated, names and contact details are protected before necessary excerpts are sent for processing. Results are retained only if you choose to create a share link, and those links expire automatically after 7 days.

Full details are available in our privacy policy.

🙋 Frequently asked questions

Basic analysis is free and requires no account. Optional premium reports use a one-time payment. Statistics are computed in your browser and your original file is not stored.