Premium Historic Tweets Search



Hello guys, I am new to this forum and the Premium access for developers.
I got a bit lost in all the info regarding this service. Could anyone please direct me to a good starting guide regarding how to fetch historical Tweets please or provide a sample code for how to search for some historical Tweets with containing a certain keyword? I mainly work with Python.
Thanks a lot!


Take a look at our search-tweets Python module. It will work with both the premium 30-day, and full archive search endpoints (assuming you have access). We’ve also got a blog series with some Python-based deep dives on using it.


In addition to Andy’s tools and blog posts, you can find some good information about the Premium tier of products here:


@andypiper is it possible to use the 30-day premium for time series analysis of hashtags, tweets, and mentions?


You can definitely use the 30-day premium product for a time-series analysis, albeit it will be limited to just the previous 30-days (obviously).


@LeBraat thanks, please suggest some materials to read until by premium access is granted. I am also eagerly waiting for my access to be granted!


It would be worthwhile to get to know our rule syntax and operators, as well as whatever python tool you end up deciding to use.

Here are a few resources to start:
Translating plain language to rules
Building a rule with Premium Operators
Search Tweets Python

Hope this helps!


Thanks for the feedbacks


Thank you guys!


Do you have any idea, how much it would cost to obtain a historical aggregated data for six months or more?
Aggregated = tweets counts for a certain hashtag or tweet.


That varies greatly depending on the hashtag or tweet.

Once you update to the paid version, you can use the counts endpoint to get a better understanding of the volume associated with a query that might include hashtags or other operators.


Do we have any Java SDK for Twitter Premium API for historical search?(fullarchive end point)
@LeBraat @andypiper

Thanks bunch in advance.
I believe, count endpoint for full archive premium api will get you the count info, before you start fetching the actual data, is that correct?


There is no Java SDK at this time, but do feel free to develop one! :slight_smile:

Yes, you’re correct that the count endpoint would provide counts ahead of fetching data. This is also useful for scenarios where you only require timeseries of data rather than the specific data itself.