Which option pricing models agree best with the market, given the asset price is known?

Assuming you can somewhat forecast the underling asset price movement, and you want to translate this value into the corresponding option price. In practice, which are the better models for this task?

Black-Scholes sometimes gives me unrealistic implied volatility, even for near the money cases, and Heston model fitting is not very stable.

I don't have a huge set of option price data to test.

• Plugging in the implied volatility is the most accurate. – SmallChess Oct 8 '15 at 23:37
• What do you mean forecast underlying asset price movement, you mean you can know $S_T$ at titme $t$? – SRKX Oct 9 '15 at 1:47
• @SRKX yes, sort of, so you may want to use option to amplify your gain here. – user2188453 Oct 9 '15 at 12:29