What Is a SAFE Note and How Does It Affect Your Cap Table

Here’s a stat that should worry you if you’re currently stacking SAFEs without tracking the math closely: Carta’s Q1 2025 State of Private Markets report found that SAFEs now account for 90% of pre-seed deals and 64% of seed deals on their platform, according to Waveup’s 2026 SAFE guide. 

That means almost every founder raising early-stage money right now is using an instrument that feels simple to sign and is genuinely easy to get wrong on the cap table side.

The reason it’s easy to get wrong isn’t the document itself. YC’s standard SAFE is five pages, compared to 15-20 pages for a typical convertible note, per Startup Project’s 2026 breakdown. 

The reason it’s easy to get wrong is that founders sign several of them over 12-18 months, never model how they stack, and then get a real surprise when the priced round finally happens and all those SAFEs convert at once.

Let’s fix that. Here’s what a SAFE actually is, how it converts, and exactly what it does to your ownership.

What a SAFE Note Actually Is

A SAFE, Simple Agreement for Future Equity, is not debt. It has no interest rate and no maturity date, which is the single biggest thing that separates it from a convertible note, per Wall Street Prep’s breakdown of SAFE mechanics. 

Y Combinator partner Carolynn Levy drafted the original SAFE in late 2013 specifically to replace the convertible note, with the stated goal of cutting the legal cost and negotiation time of a pre-seed round from weeks down to days, according to Waveup’s guide.

In plain terms, a SAFE is a promise: an investor gives you money now, and in exchange, that money converts into equity later, when you raise your next priced round (typically a Series Seed or Series A), on terms set by a valuation cap, a discount rate, or both.

Nothing converts at signing. That’s the part that trips founders up. You can raise five SAFEs over a year and your cap table won’t visibly change until the day your next priced round closes and all of them convert at once.

The 2018 Change That Founders Still Get Wrong

If you learned about SAFEs a few years ago, or you’re looking at an old template, there’s a change you need to know about. Y Combinator rewrote its SAFE in 2018, switching the entire mechanic from pre-money to post-money, per Waveup’s guide. This wasn’t a small tweak. It shifted the entire dilution burden onto founders.

Under the old pre-money SAFE, if you stacked multiple SAFEs before your priced round, the dilution from each new SAFE got spread across everyone, investors included. 

Under the current post-money SAFE, each SAFE holder’s ownership percentage is locked in at signing, based on the post-money cap, and every subsequent SAFE dilutes only the founder’s pool, not earlier SAFE holders, according to SheetVenture’s 2026 valuation cap data. 

Post-money SAFEs now represent over 80% of all SAFE instruments in the market, per the same source.

This is the single most important mechanical fact in this entire article: when you sign a post-money SAFE, you are not just agreeing to give up a percentage of the company. You are agreeing that every dollar of future SAFE money you raise before your priced round comes directly out of your own pocket, not the earlier investor’s.

Valuation Cap vs. Discount Rate

Most SAFEs use one or both of two mechanisms to determine what the investor eventually gets.

The valuation cap sets a ceiling on the price at which the SAFE converts, protecting the investor if your company’s valuation shoots up before the next priced round. 

If an investor puts in $500,000 on a $5,000,000 post-money cap, the math is direct: $500,000 divided by $5,000,000 guarantees that investor 10% ownership, regardless of what valuation your priced round actually lands at, according to Terms.Law’s 2026 SAFE calculator guide.

The discount rate gives the SAFE investor a percentage reduction on the price per share that new priced-round investors pay. A 20% discount is standard: if new investors pay $1.00 per share, your SAFE investor pays $0.80 per share, per Startup Project’s guide.

When a SAFE has both a cap and a discount, here’s the mechanic that actually decides the outcome: the calculator (and the conversion itself) compares the price per share under the cap against the price per share under the discount, and the SAFE converts at whichever price is lower, giving the investor more shares, per Terms.Law. 

SheetVenture’s worked example shows just how different the outcomes can be: at $0.80 per share, a discount-based conversion might produce 125,000 shares, while the cap-based price produces roughly double that. In the large majority of real conversions, the cap wins decisively. 

The discount only outperforms the cap when your priced round lands very close to the cap valuation itself.

What Counts as “Normal” in 2026

If you’re negotiating a cap right now, here’s where the market actually sits. According to SheetVenture’s 2026 data, sourced from Carta, PitchBook-NVCA, and AngelList:

• Median post-money SAFE cap for non-AI companies: $6M-$10M at pre-seed, $10M-$15M at seed
• AI/ML companies: command a 2-3x premium, with pre-seed caps of $12M-$25M and seed caps of $25M-$50M+
• Post-Demo-Day YC companies specifically: $15M-$25M standard, $25M-$50M+ for AI-focused startups

On the instrument mix itself, Carta data shared by investor Peter Walker and cited in Crunchbase’s coverage of uncapped SAFEs found that among SAFEs issued for U.S. startups in 2024, 61% used a cap only, 30% used a cap and a discount, 8% used a discount only, and just 1% used neither a cap nor a discount (an uncapped SAFE). That last category is worth flagging: one investor quoted in the same piece was blunt about what an uncapped SAFE signals, calling it evidence that “at least 1/12 VC funds has no idea what it’s doing.” If someone offers you an uncapped SAFE, understand that you’re being asked to give up a defined ceiling on dilution, and be clear-eyed about why.

A Worked Example: How SAFEs Actually Hit Your Cap Table

Here’s where the abstract math becomes a concrete problem. Say you raise three SAFEs over 18 months:

• SAFE 1: $250,000 on a $5M post-money cap
• SAFE 2: $400,000 on a $8M post-money cap
• SAFE 3: $350,000 on a $10M post-money cap

None of these SAFEs affect each other directly. Each investor’s ownership is calculated against their own cap, independently, at the moment of conversion. But all three convert into your priced round at the same time, and all three dilute only you and your existing shareholders, not each other. 

That’s the post-money mechanic working exactly as YC intended, and it’s also exactly why stacking SAFEs without modeling the combined effect is dangerous. You can do the math on each SAFE individually and still be blindsided by what they add up to once they all convert on the same day.

This is precisely why Startup Project’s 2026 guide includes a specific warning threshold: once you’ve stacked more than $1.5M-$2M in SAFEs, that’s generally a signal it’s time to stop and price a round, rather than adding another SAFE on top. 

Past that point, the combined dilution from a future conversion event becomes hard to reason about clearly, and you’re negotiating each new SAFE without a clear picture of what your actual ownership will look like once everything converts.

The Cap Table Mistake That Causes the Most Damage

The single most common failure mode isn’t misunderstanding any individual SAFE’s terms. It’s not modeling all outstanding SAFEs together, as a group, before agreeing to a priced round’s terms. 

A founder who has three or four SAFEs outstanding needs to model every one of them converting simultaneously, at their respective caps and discounts, against the actual priced round valuation, before signing that round’s term sheet. 

If you only model your last SAFE, or your average cap, you will be wrong, and you’ll find out how wrong on the day your cap table gets finalized, not before.

Do this modeling with a real cap table tool or your legal counsel before you’re deep in Series A term sheet negotiations, not during them. By the time a lead investor is asking pointed questions about your fully diluted ownership, you need to already know the answer cold.

The Bottom Line

A SAFE feels simple because the document is short and the negotiation is fast. The cap table consequences are not simple, and they don’t show up until your next priced round forces every outstanding SAFE to convert at once. Know whether you’re on a pre-money or post-money SAFE. 

Know your cap and your discount, and know which one will actually control the conversion math. And model every SAFE you have outstanding together, as a group, well before you’re sitting across from a Series A investor who’s already done that math for you.

If you’re stacking SAFEs and want a clear picture of exactly how they’ll hit your cap table before your next priced round, that’s exactly the kind of work we do at Stravyn Hill. Model it before you sign the next one, not after your Series A term sheet arrives.