The single number in your SAFE that decides how much of the company you keep when it converts. Most founders sign one without doing the math. The math is not hard, and once you've done it once you can never unsee it.
Reading time: 7 minutes. No legal jargon. No "consult an attorney" disclaimers (you should — but you should also understand what you're signing).
A SAFE (Simple Agreement for Future Equity) doesn't give the investor shares today. It gives them the right to shares later — usually when you raise a priced round (Series Seed or Series A). The valuation cap sets the maximum company valuation the investor's money will convert at, regardless of how much higher the priced round actually values the company.
Concretely: an investor puts in $100k on a SAFE with a $5M post-money valuation cap. You raise a Series A at $20M post-money two years later. The investor's $100k converts as if the company were worth $5M, not $20M. They get 4× more equity than the Series A investors got for the same dollar.
That's the whole mechanic. Everything else is variations on who gets to define what counts as "valuation" and what happens at the edges.
At the moment you sign a SAFE, the company has no formal valuation. You're agreeing to figure out the math later. The cap is your IOU on that math.
Lower cap = more dilution for you. If a $3M cap converts at a $30M priced round, the SAFE investor takes 10× the equity their dollars would buy at the priced round price. Stack four or five SAFEs at low caps and a chunk of your cap table is gone before you've raised your first priced round.
Higher cap = better outcome for you, harder ask for the investor. Investors push for low caps because that's where their upside lives. Your job is to defend a cap that prices the company against where you'll actually be when you raise next, not where you are today.
The mistake most pre-seed founders make: anchoring the cap to "what I think the company is worth right now" instead of "where will I be raising at in 12-18 months, minus a fair discount for the investor taking risk early."
You raise $500,000 on a post-money SAFE with a $5M post-money valuation cap.
At the cap, the SAFE converts to $500k ÷ $5M = 10% of the post-money cap table.
Eighteen months later, you raise a Series A at $25M post-money. New investors put in $5M for 20% of the company. The SAFE was supposed to convert at the priced round's terms, but the cap kicks in. Your earlier investor still gets their 10% — at the priced round's per-share price, that 10% is now worth $2.5M of equity for their original $500k. They 5×'d on paper.
Now run the same scenario with a $10M cap instead. The SAFE converts to $500k ÷ $10M = 5% of the cap table. At a $25M Series A, that 5% is worth $1.25M for their $500k. They 2.5×'d, which is still excellent for an early-stage check.
Difference for you: a 5% slice of the cap table, before any other SAFEs, options pool top-up, or Series A investor stack. Multiplied across multiple early checks, the cap you negotiate is the difference between owning 50% of your company at Series A and owning 35%.
YC standardised on the post-money SAFE in 2018. If the document says "post-money" but the cap is described as if it were the pre-money valuation, the math silently changes against you. Check the YC SAFE template version on the document — anything before October 2018 is pre-money by default.
If you raised a previous SAFE at a $6M cap and a new investor offers $4M, that's a down round in everything but name. Either the previous investor takes a haircut (rare, requires their consent) or you take more dilution than the numbers suggest.
Standard SAFEs let the investor pick the better of cap or discount at conversion. Some bespoke versions stack them — the investor gets the cap price and a 20% discount on top. That's a non-standard ask. Push back.
An MFN says: if you give a later investor better terms, this earlier investor retroactively gets those terms too. Reasonable in principle. Dangerous when it has no expiration — it can chain forward through every future SAFE you sign, rewriting your cap table years from now.
Sometimes pitched as "founder-friendly." It is — until the priced round, where the SAFE converts at whatever the priced-round price is, with no upside for the early investor. Most sophisticated investors won't sign this. If yours did, understand why before you celebrate.
Prism reads the document, extracts the cap, the discount, the MFN clause if there is one, and the SAFE template version. Then it compares against your playbook — your rules for what a fair cap looks like at your stage, what discounts you'll accept, and which clauses are non-negotiable.
The output: categorised findings (low / medium / high), market context (how this cap compares to similar-stage rounds in your sector), and ready-to-send redlines against your playbook. Not generic advice. Your rules, applied to this specific document.
If your playbook says "no MFN clause without a 12-month sunset" and the document has one without, that's a finding. If your playbook says "minimum cap = 4× last round's cap" and the new offer is 2.5×, that's a finding. The work isn't in identifying the issues — it's in defining the rules. Prism gives you both.
Closed alpha. Premium tier — €49/month, no free tier, no demos. Read the home page if you haven't, then drop your email.
SAFE discount rate — the other lever besides the cap, and which one wins at conversion.
Post-money vs pre-money SAFE — why the 2018 post-money default quietly costs you more dilution.
How a SAFE converts — where the cap actually does its work.
If there's a term you're trying to understand right now and it's not on this list, tell us — the order we write these in is driven by what founders ask for.
