WEIR

Free tool

Twitter Sentiment Analysis Tool

Type a brand, keyword or hashtag to score up to 60 recent X posts about it as positive, neutral or negative, over the last 1, 3 or 7 days, with the posts behind each label. It uses an English word list, not AI, so read the examples too.

About 63 credits at most, for up to 60 posts, from your free Weir credits (5,000 a month, no card).

The X search

iPhone -filter:nativeretweets

From 2026-11-15 17:47 UTC

How it works

The tool searches X for your keyword, brand or hashtag over the window you pick, newest first, reposts left out, and reads up to 60 posts through Weir's API. Each post is scored with an English list of positive and negative words, and a “not” just before a word flips it.

It isn't AI, so it misses sarcasm, slang and context; read the example posts before you quote a number. To keep watching what people say, turn the search into a Weir monitor that sends new posts to Discord, Slack, Telegram, a webhook or an RSS feed.

Using the result

Compare positive with negative rather than with the total, and compare brands over the same days. The share of voice calculator shows how much of the conversation each brand holds. For your own analysis, see sentiment analysis in Python and sentiment datasets; for complaints in particular, customer complaints on X.

Questions

How does Twitter sentiment analysis work?

It collects recent posts that mention a keyword and labels each one positive, negative or neutral, with a word list or a trained model, then reports the share of each. This tool uses an English word list: “love” counts as positive, and a “not” before a word flips it.

How accurate is Twitter sentiment analysis?

Good enough to show direction and big shifts, weak on single posts: sarcasm, slang, emoji and posts comparing two brands are often mislabeled. Read the example posts before quoting a number.

Why are most posts neutral?

Many posts are news, links, questions or announcements with no opinion, so neutral is usually the largest group. Compare positive with negative rather than with the total.

Can I check sentiment about a competitor?

Yes. Any public keyword, brand, @handle or hashtag works, so you can compare your brand with competitors over the same days.

Does it work for Japanese or other languages?

Not well. The word list is English, so posts in other languages mostly come out neutral. For other languages, read the posts themselves or use a model built for that language.

Is the sentiment analysis tool free?

It reads up to 3 pages of 20 posts, about 63 credits at most, from the free plan's 5,000 credits a month. No card and no X login.