Our model flags two fights on UFC Vegas 120, and neither is the main event.
The card, also billed UFC Fight Night 284, runs Saturday at the UFC APEX in Las Vegas. All 12 bouts have real betting-market odds behind them.
Alexia Thainara at 75.5% over Amanda Lemos is our strongest read, most likely by decision. Ty Miller at 73.2% over Billy Ray Goff is the other, most likely by KO/TKO. Both sit in the HIGH band. Nothing on the card reached LOCK.
A note on names before the rest of the card: there are two Millers here. Ty Miller is flagged. Juliana Miller, who meets Ravena Oliveira, is not.
In the main event, our model has Quillan Salkilld at 65.9% over Mateusz Gamrot, by decision. That number is not flagged, which is our way of saying the model sees a lean and not much more.
Juliana Miller comes in at 65.8% over Oliveira, by decision.
Yadier Del Valle is 64.1% over Darren Elkins, by decision. Elkins is fighting for the last time; more on that elsewhere. The model does not weight goodbyes.
Manoel Sousa is 61.9% over Richie Miranda, by KO/TKO. Steven Asplund is 60.0% over Guilherme Pat, by decision. Carol Foro is 59.6% over Gigi Canuto, by decision.
Diego Ferreira is 58.3% over Billy Quarantillo, by decision. Diyar Nurgozhay is 57.9% over Bruno Lopes, by KO/TKO. Jose Montanha is 55.5% over Louie Sutherland, by KO/TKO.
The closest fight on the card is Miles Johns at 53.4% over Gianni Vazquez, by decision. That is close enough to a coin flip that the number is mostly a formality.
That bout arrived late. Jessie Rosas was originally booked against Johns and withdrew. Vazquez stepped in at featherweight. He is 0-1 in the UFC, having lost a unanimous decision to Javid Basharat on Feb. 7, 2026.
The method reads split eight and four. Eight come back as decisions: Salkilld, Juliana Miller, Del Valle, Asplund, Foro, Ferreira, Thainara and Johns. Four project as KO/TKO: Sousa, Nurgozhay, Montanha and Ty Miller.
What that pattern says about the card is modest. Method calls are the softer half of what the model does, and last weekend in Belgrade they came in at 3 of 14 while the winner picks went 10 of 14. Two weak method cards in a row make a slump. We are not going to invent a cause for it.
The honest summary of a 12-fight APEX card is that two fights clear our bar and 10 do not. The unflagged picks still carry a number, and readers can see all of them above, but a 53.4% and a 75.5% are different objects and we would rather label them plainly than dress up the middle of the card.
Nothing here is a wagering recommendation. The numbers are the model's, published before the fights.