How Seedcase works.
From a simple idea to mathematical certainty. Here's the full picture.
Curve of degree 2, as a threshold of 3 requires. Any three shares redraw it and reveal the secret at f(255); any two leave it undetermined.
Shamir's Secret Sharing.
In 1979, cryptographer Adi Shamir solved a fundamental problem: how do you split a secret so that no single piece reveals anything, but enough pieces together reveal everything? His answer was a mathematical scheme based on polynomial interpolation. Any M fragments reconstruct the original. Any fewer and the secret remains completely out of reach, not by luck, but by mathematics.
Seedcase uses SLIP-39, a standard created by SatoshiLabs, also implemented by hardware wallets like Trezor. Each fragment becomes a sequence of words, easy to write down or store on any support.
We don't roll our own cryptography. SLIP-39 has been audited and reviewed by the security community for years.
Read the SLIP-39 specificationYou decide the threshold.
Choose how many fragments to create and how many are needed to reconstruct. Every threshold is mathematically valid.
Example of a 3-of-5 scheme. Five fragments are created, only three are needed to reconstruct the seed phrase.
Personal use
Three fragments, any two reconstruct. Simple, resilient, ideal for everyday peace of mind.
Family redundancy
Five fragments distributed across trusted locations and people. Survives multiple losses without compromising security.
Institutional
Seven fragments, five required. Maximum redundancy for high-value or shared-control scenarios.
A second lock on every fragment.
Each fragment can be encrypted with a password before storage. Even someone who gathers enough fragments to reconstruct your seed phrase gets nothing without it.
We never store it. It doesn't exist anywhere except where you put it. The threshold scheme and the password protection operate independently: breaking one doesn't compromise the other.
Password protection is part of the SLIP-39 standard.
Organize fragments into groups.
Real-world trust isn't flat. You may trust your closest family fully, your siblings partially, and a notary professionally. You want each circle to play a different role in recovering your seed.
Seedcase lets you organize fragments into groups. Give each circle its own group, with its own threshold. In the family, any two of three are enough. Among the professionals, one of two is enough. You need both groups to recover the seed. The family cannot do it without a professional, and a professional cannot do it without the family.
Group-based reconstruction is part of the SLIP-39 standard.
Store each fragment where it makes sense.
The supports are the easy part. The allocation is what makes a scheme survive. Four rules that hold at any threshold.
Separate the places.
Two fragments under the same roof count as one. A fire, a burglary, a flood: anything that reaches one reaches the other. Spread them across locations that fail independently.
Separate the kinds of trust.
Yourself, the people close to you, an institution. A scheme leaning on a single category inherits that category's failure mode: you can be incapacitated, a family can fall out, an institution can close.
Match durability to the role.
The fragment meant to outlive the house goes on metal. The one you reach for often can live on the machine you already use. Not every fragment needs to survive the same thing.
Never let connected devices reach the threshold alone.
Your computer and your phone are online. If those two are enough to reconstruct, a remote compromise is enough too. Keep at least one required fragment offline.
Custodians, not just supports.
A fragment doesn't have to sit on an object you own. A trusted person can hold one, and so can a notary or a legal professional: a sealed fragment deposited with them is institutionally recognized and independent from any digital infrastructure. It is also useless on its own. That is what makes a notary group work: they can be required without being trusted.