SAFTs vs. SAFEs for Web3 Startups: How Pre-Token Funding Actually Works

Two Instruments, One Very Different Set of Risks

Somewhere between your first investor conversation and your first term sheet, someone will tell you to "use a SAFT." The advice usually arrives with the confidence of a settled default — as if a SAFT were simply the crypto edition of the SAFE, the same instrument with a token bolted on. It is not. The two acronyms differ by a single letter, and that letter hides a structural gap most founders don't notice until it costs them.

A SAFE — Simple Agreement for Future Equity — is the instrument Y Combinator introduced in late 2013, and it is now used by almost all YC startups and countless non-YC startups as the standard early-stage fundraising tool. An investor gives you money today and receives equity in your company when a priced round later triggers conversion. A SAFT — Simple Agreement for Future Tokens — borrows the same skeleton but converts into something else entirely: the investor buys the right to receive tokens when your network launches, a structure that first appeared in the 2017 SAFT framework.

That one-letter swap drives two questions this guide takes apart in turn:

• What actually converts? Equity in a company versus tokens on a network — two different assets, with two different bodies of law behind them.
• What is your securities exposure? The SAFE's treatment is well-settled; the SAFT's is not, and that unresolved status is where founders get caught.

The reason founders reach for the SAFT so casually is precisely that the SAFE is so familiar. The name feels like a known quantity. Treating it as one is the mistake.

What a SAFT Was Designed to Do

The Simple Agreement for Future Tokens did not emerge from a scam or a shortcut. It came out of a serious, good-faith attempt to build a compliant path for token sales at a moment when the SEC had offered founders almost no guidance. On October 2, 2017, Cooley partner Marco Santori published the SAFT Project whitepaper, "Toward a Compliant Token Sale Framework," alongside Protocol Labs founders Juan Batiz-Benet and Jesse Clayburgh. If you are evaluating pre-token funding today, it helps to understand what these lawyers were actually trying to solve.

Start with the part the drafters got right, and never disputed. The SAFT itself is a securities instrument. It is a contract sold to accredited investors for money now, in exchange for tokens delivered later, and the whitepaper says so in plain terms: "The SAFT is a security. It demands compliance with the securities laws." In practice, that meant selling the SAFT through a Reg D private placement to accredited investors — the same exemption a startup uses for a priced equity round. There was no attempt to dodge securities law at this stage.

The entire theory rested on the second half. The premise was that a SAFT funds development, and by the time the network is genuinely built and the tokens are delivered, those tokens do something — you can spend them, stake them, use them to buy storage. At that point, the argument went, the tokens are no longer investment contracts. The whitepaper stated it directly: "The resulting tokens, however, are already functional, and need not be securities under the Howey test." A regulated instrument at the front end, a plain utility token at the back end.

The marquee deployment was Filecoin. In September 2017, Protocol Labs completed a SAFT sale that raised over $205 million from more than 2,100 accredited investors — accredited-only, precisely because the front-end instrument was a security. It was, at the time, one of the largest token raises ever conducted, and it was structured to be defensible.

Hold onto that structure, because it is the hinge of this entire discussion. The front end of the SAFT was never seriously contested. The back end — the assumption that functionality alone strips a token of its securities character — is where the model met resistance, and it is why the SAFT is no longer the default answer for web3 founders raising against a future token.

How That Theory Held Up: Not Well

The SAFT's whole pitch rested on a timing argument: sell investors a contractual right today, deliver a functioning token later, and by the time the token exists it's a usable good rather than a security. Two federal courts in the Southern District of New York took that argument apart in 2020, and they did it the same way — by refusing to look at the SAFT in isolation.

In SEC v. Telegram Group Inc., the court granted a preliminary injunction blocking Telegram from distributing its "Gram" tokens, which it had funded through a SAFT pre-sale that raised roughly $1.7 billion from initial purchasers. The key move was analytical: the court treated the initial private SAFT sale and the anticipated resale of Grams into the secondary market as parts of a single scheme to distribute unregistered securities. Applied to the whole arrangement rather than the paper contract alone, the Howey economic-reality test was satisfied — the SAFT couldn't launder away the investment character of what was really going on.

SEC v. Kik Interactive Inc. reached the same destination by a slightly different road. The court granted summary judgment for the SEC, holding that Kik's SAFT pre-sale to accredited investors and its later public sale of "Kin" tokens were part of a single integrated offering. Because the two sales collapsed into one, the accredited-investor pre-sale lost the Regulation D private-placement exemption it had been counting on, and Kik was found to have violated Section 5 of the Securities Act. The lesson from both cases is the same: courts will follow the money through the entire distribution, and a clever contract structure at the front end does not immunize the offering at the back end.

That distinction matters for how you plan. Telegram and Kik remain the case-law backbone for how a SAFT gets analyzed, and any structure you build today should be able to survive their reasoning even if you never draw the SEC's attention. A softer enforcement posture lowers the odds that a regulator comes knocking; it does nothing to change what a court would conclude if one did.

SAFT vs. SAFE: What Actually Converts

The clearest way to see the difference between these two instruments is to ask a single question: when the paper converts, what do you actually end up holding? That answer tells you almost everything about the legal risk you are taking on. On one side you have decades of settled corporate law. On the other you have the unresolved securities question you read about in the last section.

A SAFE converts into equity — shares of your company — at a future priced round, usually a Series Seed or Series A. A convertible note does the same thing, except it starts life as debt that converts into equity on a triggering event. A SAFT, by contrast, was designed as "an investment contract whereby investors purchase the right to receive tokens in the subsequent network launch." It converts into tokens, not shares.

Line the two up on the questions that actually matter and the contrast is stark:

• What converts: A SAFE or note converts into equity. A SAFT converts into tokens that may not exist yet.
• Asset-class certainty: Equity is a settled, well-understood asset class with a deep body of law behind it. A token's legal character is far less settled.
• Securities posture: Equity issued to investors is straightforwardly a security, and the rules for selling it privately are mature. A token's securities status is contested — the exact fight the SEC pressed in Telegram and Kik.

Here is the practical upshot for a company whose token does not yet exist. Converting an investment into equity is a solved problem — the instruments, the exemptions, and the case law are all in place. Converting into tokens imports the entire Howey fight before you have written a line of network code. For that reason, for many pre-token companies a SAFE or a convertible note is simply the cleaner path, even when a token is squarely in the long-term plan.

The Modern Approach: Equity Plus a Token Warrant

After the SEC shut down the standalone SAFT, the market did not abandon pre-token fundraising. It split the deal in two. Instead of selling one instrument that bundles capital and future tokens together, founders now separate the equity raise from the token rights and document each as its own thing.

The structure works like this. First, you raise capital into your company using a settled equity instrument — typically a SAFE or a convertible note, the same instruments Web2 startups use. That piece funds the business and converts to equity, with no tokens attached. Second, and separately, you grant investors a token warrant or a token side letter that gives them a contractual right to a future token allocation if and when your network issues one. Founders raising equity via a SAFE increasingly use exactly this pairing, documenting the equity raise and the token rights as two distinct instruments rather than one hybrid.

The point of the split is decoupling. Your equity raise is a settled, well-understood securities transaction that closes today. The token allocation is a contingent, forward-looking right that may never mature — and if it does, it does so against the facts that exist at that later date. Keeping them apart means a problem with the token side does not automatically unwind the money already in your bank account.

On the equity side, a compliant private raise in the United States most often runs through Regulation D, Rule 506(c). That exemption permits general solicitation and advertising — you can talk about the raise publicly — but only if every purchaser is an accredited investor, the issuer takes reasonable steps to verify each purchaser's accredited status, and the other conditions of Regulation D are met. You also file a Form D with the SEC within 15 days after the first sale. In plain terms: you can market the round openly, but you cannot cut corners on who buys in or on proving they qualified.

Treat the version above as the prevailing shape of a 2026 pre-token raise, not a template you can run on your own. Whether 506(c) is the right exemption, how the token warrant should be drafted, and how the eventual distribution gets structured all turn on your specific facts and demand your own counsel.

When a SAFT Makes Sense — and When It Just Adds Risk

A SAFT is not dead on arrival. In narrow situations it can still be the right instrument — the problem is that most pre-token companies reach for it out of habit rather than fit, and in doing so they import unresolved securities risk that a cleaner structure would sidestep. Before you paper a SAFT, work through a short set of questions honestly. If you cannot answer yes to each one, you are probably taking on exposure you do not need.

• Is the token genuinely central to how the product works, rather than an afterthought bolted onto an equity company?
• Is the network near-functional — close to launch — rather than years of engineering away, which was the exact gap the Telegram and Kik courts seized on?
• Are your investors sophisticated and accredited, so a Regulation D private placement is realistic?
• Are you prepared for full securities compliance and honest disclosures — no promises or implications that the token will appreciate, which is what the SAFT whitepaper assumed away and courts refused to accept?

For most readers, the answer to at least one of those is no, and the equity-plus-token-warrant structure covered earlier is usually the cleaner path. It lets you raise on a familiar instrument today and hand investors token upside later, without asking a court to bless a theory the SEC has already litigated against. Where the token is core, the network is close, and your cap table is genuinely institutional, a SAFT can work — but that is a determination to make with securities counsel, not from a template. The choice of instrument is a means, not a strategy. Pick the structure that survives a court's reasoning, not just a soft enforcement year.