I'm sure you'll get a million different responses, but there is no right way per se. If you want something that you need to stand up to the scrutiny of the IRS then a 409A valuation may be required ($1K-$2K)... it'll be quite a small number unless you've got meaningful revenue. Would generally avoid risk-adjusted DCF as it's kind of indefensible in most cases. If you have a solid product and an *awesome* engineering team then you can probably create an argument for $1m/engineer if you think you'll be bought.
Comps are generally the way to go. Lots of published data on early-stage valuations and hard to argue with. Even then, you probably want to discount that significantly to account for risk and the fact that those valuations are based on new money taking preferred (which is typically valued 4X-10X of common in early stage companies).
Ultimately the number is going to be what you want it to be. So identify the number and then rationalize it however you want. The only practical impact now is that the higher it is, the more expensive equity/options are for early employees.