Find historical Twitter Datasets, your ultimate source for research
Find the conversations that matter and skip the noise.
Book a meeting with stu@texifter.com to find out more.
Precisely targeted X data for researchers: see the cost first, then analyze it in DiscoverText.
Collecting social data for research is slow and unpredictable. You rarely know how many items a search will return, or what it will cost, until you’ve already spent the credits.
Query syntax is unforgiving. One misplaced parenthesis, a missing operator, or a badly grouped OR can quietly change a search. You can end up with a dataset that misses what you wanted and includes what you didn’t, and you may not notice until the analysis is underway.
The data still has to go somewhere. Even a good search leaves you with exports to clean, format, and load into analysis tools. This is more time spent before the research starts. Collecting from X outside its terms can put your data, and your publication, at risk. Many researchers aren’t sure whether their method is compliant.
XSifter is an X-compliant Enterprise application to license specific historical X/Twitter datasets or collect data in realtime that will be stored and analyzed on the DiscoverText research platform. Whether it’s a few thousand Tweets once, or a million Tweets a month, XSifter enables you to retrieve the exact information required for your research.
Please e-mail stu@xsifter.com if you cannot find an answer to your question.
We have launched as of September 1, 2026. For a demo, book a meeting with the founder. For now, the USD pricing model is:
- $100 to license the first 5,000 posts and take a free training session.
- $15/1,000 to license additional posts for academics or non-profits.
- $25/1,000 to license additional posts for commercial users.
- 5% discount if you license 100,000 posts or 10% if you license 500,000.
XSifter is approved to support market and academic research projects. You can learn more about prohibited use cases targeting sensitive categories. For 14 years we have refined our approach to word sense disambiguation. With XSifter, we will create an AI-enhanced workflow to generate better query language. The original SIFTER vision in 2012 was to be able to Search Information for the Exact Result. XSifter will ensure the data gathered is highly relevant with minimal waste.
Read our software reviews and user feedback, or watch some of the DiscoverText legacy explainer videos. We have supported hundreds of peer reviewed published academic studies as well as classroom exercises since 2012. The core pillars of DiscoverText are search, filtering, deduplication, clustering, crowd source human annotation, measurement of inter-rater reliability, adjudication, a patented method for ranking human annotators, machine-learning, all delivered in a graphical user interface. DiscoverText is an X-compliant application that features the X/Twitter display.