Trezor is a line of hardware wallets developed by the Czech company SatoshiLabs that stores the private keys of cryptocurrency accounts on a dedicated device and signs transactions offline, so the keys never touch an internet-connected computer. In the Crypto & blockchain field, and specifically in wallets and custody, Trezor is a reference example of self-custody: the owner, not an exchange or bank, controls the keys and therefore the funds.
Launched in 2014 as the first commercial hardware wallet, Trezor established conventions that most of the industry later adopted, including the recovery seed written on paper, on-device confirmation of every payment and open-source firmware that independent researchers can audit.
How a Trezor device protects private keys
A Trezor protects funds by separating key storage from the networked computer that builds and broadcasts transactions. The companion software, usually Trezor Suite, prepares an unsigned transaction and sends it over USB to the device, which shows the recipient address and amount on its own screen. Only after the owner confirms with a button or touch does the device sign the transaction and return the signature.
The keys themselves derive from a BIP-39 recovery seed of 12, 20, 24 or more words generated by the device’s random number generator. Using hierarchical deterministic derivation under BIP-32 and BIP-44, one seed produces separate accounts and addresses for many blockchains. Losing the device is therefore not fatal; losing the seed is.
The following table summarizes the main security layers and what each one defends against.
| Layer | What it does | Threat it addresses |
|---|---|---|
| PIN | Unlocks the device; wipes it after repeated wrong entries | Casual theft of the device |
| Recovery seed | Restores all keys on a new device | Loss, damage or failure of hardware |
| Passphrase | Adds a user-chosen word that creates a hidden wallet | Physical seed extraction, coercion |
| Trusted display | Shows address and amount for confirmation | Malware that swaps addresses |
| Secure element | Stores secrets in tamper-resistant chip on newer models | Laboratory attacks on the chip |
Trezor models and their differences
Trezor models differ mainly in input method, display, coin support and physical hardening. The original Trezor One uses two buttons and a small monochrome screen, while the Model T introduced a color touchscreen. The Safe 3 and Safe 5 added a certified secure element, which closes a class of physical attacks that affected earlier devices.
| Model | Input and screen | Secure element | Shamir backup |
|---|---|---|---|
| Trezor One | Two buttons, monochrome | No | No |
| Model T | Color touchscreen | No | Yes |
| Safe 3 | Two buttons, monochrome | Yes, EAL6+ certified | Yes |
| Safe 5 | Color touchscreen with haptic feedback | Yes, EAL6+ certified | Yes |
Shamir backup, standardized as SLIP-39, splits the recovery secret into several shares and requires a set threshold, such as any 3 of 5 shares, to rebuild it. This removes the single point of failure of one paper seed, at the cost of more complex storage logistics.
Trezor compared with other wallet types
Trezor sits at the cold wallet end of the custody spectrum, but it differs from neighboring options in where trust is placed. Its main competitor, Ledger, relies on a secure element running closed-source firmware, whereas Trezor publishes its firmware and hardware designs so that anyone can verify the code. Both approaches are defensible: auditability versus chip-level secrecy is the core design trade-off.
| Option | Who holds the keys | Main residual risk |
|---|---|---|
| Trezor hardware wallet | Owner, on an offline device | Lost or exposed recovery seed |
| Hot wallet such as MetaMask | Owner, on an online device | Malware and phishing |
| Custodial wallet | Exchange or custodian | Insolvency, freezing, hacks |
| Multisig wallet | Several key holders | Coordination errors, setup complexity |
| Paper wallet | Owner, on printed paper | Insecure generation, single-use exposure |
A Trezor can also act as one signer in a multisig setup, combined with devices from other manufacturers. This pattern spreads supply-chain and firmware risk across vendors and is common among individuals holding large amounts and among small businesses.
Setting up and using a Trezor safely
Safe use of a Trezor depends far more on procedure than on hardware. The device should be bought directly from the manufacturer or an authorized reseller, checked for intact packaging, and initialized from scratch so that it generates a fresh seed. A device that arrives with a pre-printed seed card is always a scam.
| Situation | Recommended choice | Reason |
|---|---|---|
| Small balance, frequent use | Standard 24-word seed with PIN | Simple recovery, low overhead |
| Large long-term holdings | Seed plus strong passphrase | Seed theft alone does not expose funds |
| Shared family or company funds | Shamir backup or multisig | No single person holds full control |
| Older model without secure element | Always enable a passphrase | Mitigates physical seed extraction |
Before sending a large payment, owners should send a small test amount and verify the receiving address on the device screen, not only on the computer monitor. Firmware updates should be installed only through the official software, which verifies the firmware signature before installation.
Limitations and common mistakes
A Trezor cannot protect a user who approves a malicious transaction. If a phishing site persuades the owner to sign an unlimited token approval or to type the seed into a web form, the device has done its job and the funds are still lost. Most real losses come from social engineering, not from broken cryptography.
- Photographing or typing the seed into any computer or cloud service defeats the purpose of cold storage.
- Fake support contacts and counterfeit Trezor Suite downloads target new owners; genuine support never asks for the seed.
- Blind signing of smart contract data without readable details exposes users of decentralized applications.
- Forgotten passphrases cannot be recovered, because the passphrase is never stored on the device.
- Older models without a secure element are vulnerable to voltage glitching attacks if an attacker gains physical access.
The simple explanation
A Trezor is a small, pocket-sized safe for crypto keys. The keys that prove you own your coins stay inside the device, and the device signs payments for you after you press a button on it. Your computer never sees the keys, so even an infected laptop cannot quietly empty your account.
When you set it up, the Trezor gives you a list of words. That list is the master copy of your wallet. If the device breaks or gets lost, you can buy a new one, type in the words and get everything back. Anyone else who finds those words can do the same, so keeping the word list secret and safe matters most.
How it is used in practice
A freelance software developer who is paid partly in bitcoin keeps about two years of savings on a Trezor Safe 5. She stores the device in a desk drawer and splits the recovery secret with Shamir backup into three shares, two needed for recovery: one in a home safe, one in a bank deposit box and one with a trusted relative in another city.
Each month she moves her earnings from an exchange to her Trezor. She checks the deposit address on the device screen, sends a small test amount first and then the remainder. When her laptop is later infected with clipboard malware that replaces copied addresses, the mismatch shows on the Trezor display, she rejects the transaction, and her savings remain untouched.