I found a very interesting gambit in the sicilian defense (+ traps!!)

It all started from me trying to analyze this line as white:

The point is to eliminate the d5 response by black (as otherwise he would lose a tempo) and make the move order more precise and easier to learn. After using the opening explorer and engine, I stumbled on a gambit which is actually good for white (+0.6) and has a high chance of occurence as it follows the most natural moves.

I hope you enjoyed a lesson on the Basman-Palatnik gambit, one which you will actually get to play more often than you think as if black plays Nc6 on move 4, the gambit is already in place. Engine thinks either e6 or g6 and black is completely fine. 

Just want to point out that 9. Ne5 is probably a better option for black

My engine believes that Nd8 is the better move but not by too much. Ne5 certainly looks like a more natural move. However, the engine still gives advantage to white here.

Yes but black wins another pawn, which puts white under more pressure, because if he doesn't achieve anything quickly, black will just be 3 pawns up

What i mean is this line

4...Nc6 and 6...Nxe4 is too risky for black.

Theory tells white has too much compensations, engines agree.

Of course black has to play for "normal" chess playing alternatives at move 4 : g6, Bd7, e6, e5, Nbd7, whatever...

Engine seems to think white has a small advantage even there, but I definitely wouldn't play that as a human. However, white doesn't need to play Nxe5. White can simply castle, and after 10...Nxf3 11.Bxf3 white is only a pawn down and black hasn't even begun to castle.

Here the engine thinks that black shouldn't play 11...Qxc3 as white then plays Bd2 and Rc1 with tempi and black has no way of castling. Even without Qxc3, the engine thinks black should castle artificially (11...g6 12.Qd4 f6) with equality. I'd still take this position as it's only a one pawn deficit.

Yeah, black is probably completely fine there and 4...Nc6 seems to be innacurate, although that's not immediately obvious to me. Might have to do with the fact that I'm getting 4...Nc6 in blitz games and people don't have time to think this through.

I personally play 4. g6, and i don't think white gets much after that, but the lines after 4. Nc6 are very interesting

Right but I wasn't aware that the delayed alapin was a common opening at all. There's also no possible transition as after 1.e4 c5 2.c3 Nf6, 3.e5 is the move for white. So that must mean that this opening isn't very rare at all.

I just thought it was an interesting concept which avoids the d5 setup by black.

I think 4...Bd7 intending Qc2 Qc7! = is best

Very interesting, i found a this link to game between GM Antipov vs. Van Foreest that follows similar lines for the first 9 moves and shows a possible different response by black: 

https://thechessworld.com/articles/openings/antisicilian-for-white-delaying-the-alapin/