Only the good stuff, picked for you.
Where to type a question. Ranked on the quality of the first page and on what the company does with your queries, which is the second half of the decision most people skip.
Still the best index and still the least accountable one. The reason to keep it is that everything else is measured against it; the reason to add a second engine is that you stop seeing what it chooses not to show.
The one that matters for a second reason: it is the index behind most of the assistant products, so a page that does not rank here does not get cited there either.
The default privacy answer, and good enough for most everyday queries. The trade is that its results are thinner on obscure technical questions.
Runs its own index rather than reselling somebody else, which is rare enough to be the whole reason to care. Independent crawler, independent ranking.
Paid, no ads, and it shows: you can weight sources, pin results and block domains permanently. The subscription is the business model, which is the honest version of the deal.
Google results with the tracking taken out, proxied rather than scraped. The pragmatic choice when you want that index without that profile.
An independent British crawler that has been running longer than most of the alternatives. Results are rougher; the independence is real.
Deliberately excludes the pages that optimise for search engines, which makes it the one engine where small personal sites still surface. Useless for shopping, excellent for curiosity.
A metasearch server you host yourself, aggregating the others. The reason to run it is that nobody sees the query log but you.
Search that pays for trees, with the receipts published. The results come from Bing, which is worth knowing before you switch.
The European answer, built around not profiling users. Its own index is partial, so expect a hybrid of sources.
Unavoidable if you are searching in Chinese, and noticeably weaker the moment you are not. The results are shaped by policy in ways a foreign index is not.
The strongest index for Russian-language material, and the one whose ownership makes it a hard recommendation for anything else.
The index of record for academic work, and the citation graph is the part that is actually hard to replace.
The medical literature index, with abstracts and no interpretation layered on top. Slower to use than a general engine and the right tool when being wrong is expensive.
Computes rather than retrieves, so it answers unit conversions, integrals and date arithmetic exactly instead of pointing at a page about them.
The fastest way to learn how something is actually done, because you are reading working code rather than someone’s write-up of it.
Still where the specific error message lives. Answers age badly, so read the dates before you trust the accepted one.
The record of what a page used to say, which is the only way to check whether something changed or whether you misremembered it.
Not an engine, but the fastest way to get the vocabulary right before searching properly. Its own citations are where the real sources are.
The ranking is ours. Anything on this page that can be checked is checked against a named source, and the date we last opened that source is printed with it. If a figure and our verdict disagree, trust the figure and follow the link.
No site is listed twice. A listing means we used it for a month first, and a free tier only counts as free if it survives the trial period.