I’m conducting a registered study on the accuracy of political pundits’ predictions, using their posts on X (Twitter) as data. The project is registered on OSF: https://osf.io/s9c3x.

I need to collect tweets from about 40–60 specific accounts spanning 2020–2025. Manual collection isn’t feasible, and my attempts using tweepy, snscrape, nitter-scraper, and Playwright haven’t worked reliably.

Are there any current tools, APIs, or workflows that can handle large-scale historical tweet collection from multiple handles? Would applying for an academic API still be a viable option?

Any guidance or examples would be greatly appreciated.