This tool searches the OCR text of digitized newspapers in the Chronicling America collection through the Library of Congress's loc.gov JSON API. Each result is trimmed to the text from the first to the last matching word on the page, so you can skim a lot of newspaper pages without opening the scans.
The LOC API is often slow (a search can take anywhere from several seconds to a minute) and is rate-limited. This is a limit of the LOC server, not of this page, so please be patient.