It depends on the criteria. If you're just talking about raw peak strength, it'd be Babylon for me. Babylon can crush the game regardless of what victory condition it pursues (only considering science, culture and domination because I find the rest uninteresting). With Babylon, you can rush pretty much any unit that you can unlock through a tech you can boost without a spy and without unlocking a late-game civic. Crossbowman, man-at-arm, musketeer, biplane and more fall under this category. There are many viable domination strategies based on rushing these units. You can also build two early campuses and the Great Library, which gives you a free eureka when another player recruits a scientist. This, along with pillaging (for science mostly) your opponents with fast and powerful units (e.g. cavalry), can help you overcome Babylon's -50% science per turn penalty (pillaging is not subject to this penalty), and allow you to win a science victory probably faster than with any other civ. You need a bit more luck if you want to pursue a quick culture victory as Babylon since Babylon doesn't have any strong inherent tourism bonus. If city states like Caguana (probably the best for this strategy) and Rapa Nui are in the game, however, you can suze them (made easier with the extra envoys you get from building the first of a district type), and rush Flight to start generating tourism from the unique improvements the city states provide.
However, when I evaluate different civs' strength, I like to also take into account other factors like how dependent a civ is on map settings and luck and how difficult the civ is to play. Babylon loses a lot of points in both of these criteria because it's very dependent on its spawn and certain events like being able to kill 3 barbarian units for Bronze Working and because efficiently stringing together eurekas for maximum performance requires a lot of planning and attention to detail.
When I take all of these factors into account, Russia, Ludwig's Germany and Babylon would be my top three. Russia has a very straightforward game plan. Get Dance of the Aurora as the pantheon. Build a Lavra in the tundra in your first two cities. Build a shrine in each city after that. Found a religion with Work Ethic. Expand quickly using Russia's high production and faith. I'd say up to your third city should be in the tundra where you can build a high-adjacency Lavra. You want to build every subsequent city in a "normal" location, preferrably where there's plenty of food so you can build districts you want. If there's a good location for a Lavra, build it, otherwise build a theatre square and an amphitheatre first. With Russia, you want to hoard every single writer until the game ends and make sure you have enough slots for new writings each time you recruit a new writer. This is the most effective strategy to win a fast culture victory. On normal speed and deity difficulty, without any game modes and with just the base BBS settings, I was able to achieve 159 turns.
The reason why hoarding great writers is such a strong culture strategy is that the amount of tourism you need to generate to win the game depends on the maximum amount of culture your opponents have accumulated throughout the game. For every 100 culture your top-culture opponent generates, you need to generate a little over 200 tourism toward that and every other opponent to not get any further away from reaching the victory condition than you currently are. Each writing provides +2 culture and +2 tourism per turn (multiply both by 2 for writings by some writers like Valmiki and Rumi). In the hands of the top-culture opponent, each writer provides ~5.6 culture per turn (= 2 writings * 2 base culture * 1.4 deity difficulty yield bonus), and typically hoarding great works is how an AI opponent ends up with high culture. By preventing your opponent from recruiting a writer, you effectively generate over 11.2 tourism per turn (= 5.6 * 2) from the extra culture you're preventing your opponent from generating plus actual tourism per turn provided by the writing. With all the bonuses you can stack (+100% Printing and +100% from Pingala), one writer can generate 12 actual tourism per turn (= 2 writings * 2 base tourism * 3 from bonuses). In total, the maximum effective tourism from one writer is ~23.2 per turn (46.4 for Valmiki), which is around what you'd generate from 4 wonders by the time you can beat the game as Russia. Greece is another civ that can effectively hoard writers, but it's a bit weaker than Russia. Pericles has trouble expanding early to be able to build many Acropolises, although that can be somewhat offset by his culture-from-suzerainty bonus. One of the things having high culture allows you to do is to efficiently rotate your policies, and this synergizes very well with Greece's extra wildcard policy slot. With Pericles, I was able to recently match my personal record of 159 turns that I set with Peter, although that required me to touch the map settings a bit more (Barbarian Clans on, maximum of 18 city states placed, etc.). Gorgo's Hoplite rush is the earliest effective domination push I know of, so she can definitely expand very quickly, but it's tricky to make sure you're investing the right amount into war while not neglecting building the infrastructure you need for generating early tourism.
Most other culture civs need to unlock Flight to be able to start generating a significant amount of tourism, which is why I don't consider Khmer and Ethiopia to be on the same level as Russia and Greece. Ludwig is unique in that a bulk of his tourism is unlocked with Castles, which actually happens earlier than Printing. The culture he generates from unfinished wonders is a huge boon early game as well, and it's also something that can directly quicken the unlocking of Castles, which is boosted by adopting a tier-2 government. Ludwig is not quite as strong as Russia and Greece but is very easy to play and arguably less reliant on luck. When you're playing as Russia or Greece, you hoping that the other civ isn't in the game, and whoever your opponents are, they don't aggressively compete with you on recruiting writers. With Ludwig, you just need a couple floodplains next to your capital city centre to have a good game. Germany's ability to build one extra specialty district in each city and the Hansa's production yields are also not very reliant on luck.
Since the strongest civs I've listed are culture victory specialists, here are the best domination and science victory civs in the game.
Domination
Gran Colombia: Domination is all about timing, and Gran Colombia is rare in that its domination window of opportunity opens as early as it does for most civs (exceptions: Gorgo, Nubia, Scythia), but it basically never closes. As soon as you receive your first general and comandante general, your units get +10 combat strength and +2 movement, which is something I believe only Chandragupta can replicate (temporarily). It's a very reliable civ whose abilities are both effective and very easy to understand and utilize.
Basil's Byzantium: Basil makes domination easy by eliminating some major headaches. "How do I generate enough science to unlock the units I want quickly?" "How do I produce and maintain my units?" "How do I get a general I need in time?" "How do I break walls quickly?" You don't need science as Basil because his unique unit, Tagma, which is basically the only unit you need, is unlocked with a civic. You don't need much production or gold because you get free units from the Hippodrome. It's also not strictly necessary to get a general because your units are already ridiculously strong (+3 from Taxis, +10 from Crusade, +4 from Tagma adjacency), and the extra movement isn't as relevant with the cavalry unit class as it is for other classes. And of course, you can break walls with cavalry units.
Honourable mention: Hungary, Mongolia
Science
Scotland: You want to settle 7-8 cities with Scotland, a lot fewer than what I'd normally recommend (~12). This is because Scotland wants to maximize happiness, and if you have two copies of each luxury resource, they can reach all 8 cities, and Audience Chamber bonus applies to up to 7 cities because that's how many governors there are. 8 Scottish cities are more powerful than your average civ's 12 cities because of Scotland's happiness bonus and the 100% production bonus from declaring War of Liberation. For a large portion of the game, assuming you can keep all your cities ecstatic, you should be able to generate more scientist and engineer points with just 8 cities than you would as another civ with 12 cities. A lot of people underrate Scotland because they think the production bonus is so difficult to trigger that it's basically irrelevant, but as long as you scout the map early and pay attention to what's happening around the world, you should get your opportunities. Unless there's too much land on the map, the AI will typically be fighting each other, and it's not unlikely that, by the mid-game, you'll discover that there are two cities that allow you to declare the War of Liberation. Two is all you need for you to get 100% bonus production (and +2 movement to all units) for every single turn until the end of the game. Another mistake a lot of people make is that they think generating a lot of science is the most important part of an effective science victory strategy, but production tends to be more important, so this bonus is what makes Scotland the top science civ (other than Babylon) in Civ 6 for me.
Australia: Like with Scotland, starting at a certain point in the game, Australia can achieve +100% production per turn for the rest of the game. It's more difficult to trigger and maintain this bonus permanently as Australia, because you actually need to build units and create a situation where you can repeatedly liberate two cities. One way to do this is to find a situation where there are two other players on your continent. You'll likely be on one edge of the continent because of Australia's coastal starting bias. Whenever you're ready, you invade the player in the middle, take out all of their cities except for a couple that are close to the third player. Then, you negotiate peace and wait for these cities to rebel. After that, liberate one city to the founder. Typically, when you liberate a city, it stops receiving loyalty pressure from you, so this would be the only time you'll be able to liberate it, but if there's a different player nearby, they can still apply pressure and cause that city to flip again and again. If you can create this situation twice, you can maintain the 100% bonus until the rest of the game. Australia might be more difficult to play than Scotland in this regard, but it makes up for that with its supercharged campuses, commercial hubs, etc, although overall, I slightly prefer Scotland.
Honourable mention: Sejong, Portugal