The Superhero Database Classification number, or SHDB Class, is a number that represents the overall 'power' of a character. All traits of a character are used for calculating the Classification.
What it DOESN'T mean
This doesn't mean that a higher class would always beat a lower class character. But the bigger the difference in Class is, the more obvious it is who'll win in a fight.
How is this calculated
( INT^1.3 + (STR*0.5 )^2 + (SPE*0.5)^2 + DUR^1.6 + (POW + (SPS*SPL))^2 + COM^1.8 ) ^ TIER
Super Power Score and Level
Every Super Power has a score (SPS) that is used to calculate the Class. Each Super Power also has 3 levels (SPL). The level is set when connecting that Super Power to a character. The level determines the final score, of the Super Power, being used in the calculation.
Akuto Sai is ranked so highly because of his godlike omnipotence, reality manipulation, multiversal powers, and his ability to transcend the concept of existence itself. Characters with these abilities are often at the top of any power ranking, as they are virtually unbeatable in their respective universes and beyond.
Akuto Sai’s powers extend far beyond physical combat, placing him in a league of his own among the most powerful anime and manga characters.
It's interesting to see that even "only" A.I. are already capable of understanding at least that much about Akuto Sai.
All in all, I've been playing around a lot with AI like chatgpt and google gemini, testing how good on average their answers are when it comes to powerscaling and to "understanding" the very reasons for why some characters are very powerful, while other characters are not.
chatGPT is already giving very good answers most of the time, also has a surprising large knowledge of many different fictions , and usually has higher quality of answers than gemini.
tbh I've been asking chatgpt dozens of completely different questions and matchups, and was surprised that in like 9 out of 10 cases I could really only agree . it's become that good already.
I think that in the long run, the more time AI will have to learn and improve, they will definitely become superior powerscalers compared to "only humans", who can easily err and be blinded by their own biases.