lemmy.today. I like their "we aim to try to not defederate with other instances" policy, and they're geographically near me.
Kagi. Search engine that doesn't log or data-mine users; it charges a subscription fee. Does some neat things like specifically index and allow searching of the Fediverse. It works fine, but that's not really my interest: I really just don't want to have a search engine provider logging and data-mining my searches, and I'm happy to finally have an option to avoid that.
Wikipedia. Being the "store of all human knowledge" may be ambitious, but Wikipedia's been having a pretty good go at it, and has killed off most commercial encyclopedias.
Stuff that I don't use daily, but do probably have a good chance of having used in a given week:
Google Earth. There's no real alternative to this out there: it sucks in a lot of satellite and aerial imagery to let one get some degree of 3d view of much of Earth. Also convenient for measuring distances, including multi-hop trips.
Amazon. The world's largest retail selection and is available wherever you live. Twenty years back, one significant argument for living in or near a city was shopping choice. Amazon provides a much larger selection all over. Maybe for some of the younger crowd, that doesn't seem like a big deal, and it's a change that didn't happen overnight, but the change over time is pretty remarkable. I don't buy everything from them -- Walmart.com provides better delivery options for food and some other things that they sell, Monoprice.com has long been my go-to provider for computer cables (which have historically seen obscene markups at brick-and-mortar retail), and I used to use Newegg for their better product database. Aside from the constant nagging to subscribe to Amazon Prime, I'm pretty happy with them.
YouTube. It's the world's largest provider of on-demand video. Not only that, but for a lot of non-fiction stuff, it's a lot better than any commercial streaming service. I don't subscribe to their premium service, though I would if I could get a "no log, no analytics" guarantee of the sort that Kagi provides.
Maybe Tineye. Image-keyed index of images: feed it an image or URL of an image, and it will tell you where it's seen it, including the earliest time and the best-quality version of the image. It uses fuzzy matching, so it's capable of identifying similar images with certain kinds and levels of modification. There's no alternative for figuring out where some images may have come from or digging up less-overly-compressed version of images. I'm surprised that some of the image search providers -- which have to build an image index as well -- haven't provided this feature.
Stuff that I used to use daily:
Reddit. I was kind of sad when they transitioned to the new Web UI, but kept using the old one. But killing off the third-party clients was the breaking point for me.
Yahoo, then Altavista, then Google. Main search engines. Altavista in particular indexed Usenet for a while, and I believe that Google was the first search engine to introduce image search, which was nice.
Slashdot. Before Reddit. Didn't have Reddit's variety in topics and wasn't designed to scale up to what Reddit or the Threadiverse are, but it was a good forum for a while. I do prefer Markdown to Slashdot's HTML subset, though.