Build the 24 month model investors ask for.
Free templates assume SaaS margins and a thin hosting bill. This one knows better: training is R&D and inference is COGS, grants are cash but not revenue, credits run out, hardware is lumpy. A handful of drivers go in, a cash curve and a zero-cash date come out.
Tools, rent, legal, accounting, insurance, travel. The boring line that quietly doubles.
Costs creep. 2 percent monthly is honest for a team that is shipping.
The cash curve, month by month for 24 months. Teal above zero, orange below.
The monthly table
We send the runway, the zero-cash date, and the biggest cost driver, never your line items and never a name. One copy reaches the Lab so a human can pick up the thread if you want one. Nothing else leaves this page.
This is the lite Model Lab: a driver-based cash model, not a three statement audit. It is deliberately monthly, deliberately simple, and honest where templates lie: AI compute splits into R&D and COGS with a credit cliff, grants arrive as cash without pretending to be revenue, crypto treasuries can be stress tested at fair value, and spend creeps the way spend actually creeps. A model an investor believes is one where you chose the drivers and can defend them.
Want a human pressure-testing the model with you?
Choosing defensible assumptions and standing behind them in front of investors is the actual work. One email starts it.
Email meMore instruments in the Founder Lab. The raise this model implies: size it in the Runway Scan, price it in the Dilution Scan.