what is the strogest reaply to king gambit ?

This is what i think is a very strong reply to a main line of the Kings Gambit

Possibly of interest:

First Steps 1 e4 e5
https://www.everymanchess.com/downloadable/download/sample/sample_id/149/
Starting Out: Open Games by GM Glenn Flear (2010)
https://web.archive.org/web/20140626232452/http://www.chesscafe.com/text/hansen134.pdf
Playing 1.e4 e5 - A Classical Repertoire by Nikolaos Ntirlis (2016)

https://www.chess.com/blog/smurfo/book-review-playing-1e4-e5

https://chessbookreviews.wordpress.com/2017/05/06/playing-1-e4-e5-with-black/
http://www.qualitychess.co.uk/ebooks/Playing1e4e5-excerpt.pdf

Bologan's Black Weapons in the Open Games (2014)

http://www.thechessmind.net/blog/2014/12/6/a-review-of-bologans-black-weapons.html

https://chessbookreviews.wordpress.com/2017/05/06/playing-1-e4-e5-with-black/
http://www.jeremysilman.com/shop/pc/Bologans-Black-Weapons-in-the-Open-Games-76p3873.htm
The Kaufman Repertoire for Black & White by Larry Kaufman (2012)
https://web.archive.org/web/20140626221508/http://www.chesscafe.com/text/hansen162.pdf
https://www.newinchess.com/media/wysiwyg/product_pdf/955.pdf
Opening Repertoire: The Open Games with Black by Martin Lokander (2015)

https://chessbookreviews.wordpress.com/tag/martin-lokander/
https://www.newinchess.com/media/wysiwyg/product_pdf/7546.pdf

How to Beat the Open Games

www.gambitbooks.com/pdfs/How_to_Beat_the_Open_Games.pdf

you can try this bizarre and surprisingly sound countergambit. the amount of players that will know the inside and out variations with white is very small.

well, it's not that sound. definitely playable at anything below IM level (maybe even IM level), but stockfish gives the position above as +0.5 for white whereas almost any other line in the KG it is in favour of black. I absolutely agree that it's playable because humans don't play like engines, but from an objective point of view there might be better replies...

as sound as a countergambit can get for black. my engine which has the livebook feature gives the eval as ALL over the place depending on the engines at higher depths, like no kidding, 0.1 to 0.6's all after 3...f5 all with depths above 20. its that unclear a position.  The amount of variations  and ideas for both sides after 3.f5 is pretty darn mind boggling. if you study these lines, you can easily knock out a KG player once , maybe twice,  and still have interesting games with repeat play.

objectively, i have always thought 2.d5 was the best remedy.

i dont think eval at the 0.5 range is particularly useful in determining soundness. they are 0.5 positions whose complications make the eval mostly meaningless aside form pointing who has the small advantage, then they are 0.5 so positionally bad as to not give the losing side any winning chances an are practically lost and only computer play-hangs on draws. ( like say a symmetrical position with double isolated c- pawns)

A young Bobby Fischer lost in 1960 to Boris Spassky's King's Gambit and it prompted Fischer to study the opening deeply and publish an article in 1961 in the USCF's magazine, Chess Life, titled, "A Bust to the King's Gambit."

He recommended "1 e4 e5 2 f4 exf4 3 Nf3 d6!"

His article is here, in algebraic notation: http://brooklyn64.com/wp-content/uploads/2015/11/a-bust-to-the-kings-gambit.pdf

The article mostly looks at why 3...d6 is better than alternatives.

 Note, however, that in the 1963 U.S. Championships, Bobby as White played the less popular 3 Bc4, possibly because he may have known or suspected that his opponent, GM Larry Evans, was less familiar with that variation.

 rnb1k1nr/pp3ppp/2p5/3P1q2/8/6K1/PPP2bPP/RNBQ1BNR w kq - 8 13

bela v efstratios 1984

falkbeer counter gambit, surprising how often this pattern or variations arise, even against players who play kings gambit often as white.

https://www.chess.com/games/view/347061

Too bad Spassky didnt know this ! If someone had told him maybe he wouldnt have such an outstanding record with the king's gambit !  

And the great thing about 3...d6 is that it can be played against both 3.Nf3 or 3.Bc4. Often transposing. It can even be played against 3.Nc3. 3...d6 would be my choice if i played open games. But then again, i'm also a big fan of Fischer 

Sorry, this is incorrect.

Fischer played the King's Gambit from time to time, mostly in simuls. In the majority of those games where the gambit was accepted, he followed up with 3 Bc4. There's even a book on the bishop's gambit by Timothy Taylor called The Fischer King's Gambit.

I like how some people are talking about other things.

G5 is the strongest for grandmasters, but you can follow up with Fisher's Defence

I wouldn´t recommend "just any move" even after 2...exf4, unless you want to be crushed. Yes, KG is inferior for white, but only if black is accurate for 10 or 12 moves, if you don´t believe me ask Wesley So, a 2800 who couldn´t break Adhiban KG in Tata Steel 2017 (Adhiban 150 points elo below So).