Best Hardware Crypto Wallets: Full Review and Comparison

A review of hardware wallets: Ledger, Trezor, Coldcard, Keystone, BitBox, and NGRAVE. Comparison table, pros and cons, plus tips on backups and multisig.

||
Updated

Hardware wallet: a “cold vault” for private keys

The core idea: private keys stay inside the device, while transaction approval happens on the wallet screen — this reduces the risk of key theft through a PC, browser, or malicious extensions.

A hardware crypto wallet is a physical device that stores private keys inside it and signs transactions locally without exposing the keys to a computer or phone.

In “hot” wallets, operations are approved in a browser or on a smartphone. In a hardware wallet, the keys remain inside the device, and the signature is approved on its screen (with buttons or a sensor), so malware on the host does not gain access to private keys.

How a hardware crypto wallet works

  1. Private keys are generated and stored on the device.
  2. The transaction is signed inside the wallet and approved on the device screen.
  3. A signed transaction (digital signature) is sent to the network, while the private keys remain offline.

Principle: Who controls the keys controls the assets.

Main advantages

  • Suitable for long-term storage of large amounts: the risk is lower than with “hot” wallets.
  • Malware on a PC or phone cannot extract private keys from the device, but it can alter details in the interface — checking the address and amount on the screen is critical.
  • Suitable for DeFi and staking: connection through official apps or WalletConnect (connecting a wallet to a dApp via QR/session).

2025 context: more models have appeared for different connection scenarios (USB, Bluetooth, QR). The key criterion is the scenario: for frequent operations, the screen and confirmation convenience matter most; for “holding,” the threat model and backup quality matter more.

Well-known brands remain relevant with basic hygiene: offline seed phrase backup and checking operation details on the device screen.

3D illustration of a hardware crypto wallet: transaction confirmation on the screen, with a metal seed backup and USB/Bluetooth/QR icons nearby — cold storage and checking details on the device

Material updated → 2024–2025 models, USB/Bluetooth/QR scenarios, and typical risks of “blind” signing in DeFi are taken into account.

  • What to check before approval → address/network/amount on the device screen.
  • How to store a backup → practical backup methods (metal, passphrase, Shamir).

How to choose a hardware wallet: 5 criteria without fluff

Five criteria: details on the screen during signing, type of support (native or through integrations), recovery, as well as connection and price.

  1. Security (keys and access)
    • PIN and brute-force protection: delays, lockout, or wipe after N attempts.
    • Checking before signing on the device screen: address, network, amount, fee.
    • Physical access risks: Secure Element and/or passphrase reduce the impact of device theft.
  2. Usability (confirmation errors)
    • Screen criterion: operation details are displayed, not just “OK/Confirm.”
    • Address criterion: long fragments are readable, so any substitution is noticeable.
    • For DeFi, visibility of approve/spender (who receives the rights) matters, not “blind signing.”
  3. Coins and networks (type of support)
    • Support may be native in the official app or provided through integrations (MetaMask/Electrum/Sparrow), which means different limitations and different UX.
    • Compatibility is determined by the platform (PC/Android/iOS) and the specific app.
    • Practical check: a test cycle of “receive → send a small amount” shows which details appear on the screen during signing.
  4. Connection (USB, Bluetooth, QR)
    • USB is predictable; QR air-gap reduces the number of data exchange channels, but takes more time.
    • Bluetooth is convenient for a phone, but adds a communication channel; in a strict threat model, it is more often used as an optional mode.
    • Platform nuances: Android (OTG), iOS (model and app limitations).
  5. Price (paying for UX, not for “magic”)
    • Usually ~$50–$300+: the price increase is more often tied to the screen, touch controls, and comfort.
    • Supply-chain risk is lower when buying from official sellers; “second-hand” and “pre-configured” devices increase the chance of compromise.
    • Backup-related costs: metal, a second set, separate storage.

What else to consider when choosing

  • Software: Ledger Live / Trezor Suite / BitBoxApp and compatibility with third-party wallets for the required networks.
  • Documentation and updates: clear guides and release frequency for the required platform.
  • Backup: 24 words / microSD / Shamir (if available) and a one-time dry-run recovery test.
  • Response to vulnerabilities: speed, transparency, patches.

A typical failure is linked not to “chip hacking,” but to handling mistakes: buying from unofficial sellers, digital copies of the seed phrase, and signing without checking the address/network/amount on the device screen.

📌 Related topic: seed phrase and backups
Storage checklist: metal, separate placement of copies, passphrase, and a recovery test.

Hardware wallet comparison 2024–2025: what the money actually buys

The table is not a “ranking.” It helps match a scenario: universal multicurrency, QR air-gap, or BTC-only for strict isolation.

How to read it: 1) screen (address/network/amount during signing), 2) connection (USB/BT/QR), 3) “# of coins.” “Support” may be native (official app) or through integrations (MetaMask/Electrum/Sparrow).

Prices and number of coins are approximate: they depend on region, supply, and updates.

Model Price # of Coins Compatibility Features Advantages Disadvantages
Trezor One ~$65
(59 €)
1000 PC
Android (OTG)
OLED; 2 buttons
no SE; open-source
Low price: lower “entry threshold”
easy start and setup
open-source code
No BT or iOS
small screen: addresses are harder to verify
without SE, passphrase plays a larger role under physical access
Trezor Model T ~$219 1800+ PC
Android (OTG)
1.54″ touch (color)
microSD (Shamir)
open-source
Touchscreen: easier to verify data
Shamir Backup for backup distribution
includes Monero
More expensive than many alternatives
no BT/iOS
without SE, passphrase discipline matters more
Ledger Nano X $149 5500+ PC
Android/iOS (BT)
OLED; 2 buttons
BT + battery
SE CC EAL5+
Broad asset and integration coverage
mobile scenario: BT + battery
SE chip as a plus against physical access
Closed firmware (different trust model)
reputational disputes around brand functions/incidents
on iOS, some scenarios may be limited
Ledger Stax $279 5500+ PC
Android/iOS (BT)
E-Ink 3.7″ (touch)
Qi charging
SE EAL6+
Large E-Ink: fewer “blind” approvals
large screen and touch controls
SE EAL6+
Very high price
non-removable battery
closed firmware
Coldcard Mk4 $150 BTC-only PC
PSBT: microSD/USB
OLED + keyboard
2× SE; no BT
NFC (optional)
BTC-only: strict focus
PSBT/microSD for offline signing
multisig and protective modes
BTC only
higher entry threshold (PSBT, multisig, Sparrow)
less “daily-use” comfort
Keystone 3 Pro $149 5500+ Standalone (QR)
Android / iOS
4″ screen; 3× SE
fingerprint; no USB/BT
self-destruct
QR air-gap: no cable/BT data channel
large screen: easier to verify details
convenient for mobile DeFi via QR
QR process is slower than USB/BT
device is larger than “flash-drive” types
UX depends on apps/pairings
BitBox02 $120 ~1500 PC
Android (USB-C)
Compact USB-C
touch edges
microSD backup
microSD backup: fast recovery
balance of “transparency + hardware”
BTC-only version available
No iOS on the standard model
fewer networks than Ledger
touch controls require adjustment
NGRAVE ZERO $398
(398 €)
1000+ Standalone (QR)
Android / iOS
4″ touch; camera
SE EAL7; biometrics
fully offline
Focus on “hardware” and physical risks
large screen + offline approvals
Graphene backup as a separate scenario
Very high price
less mainstream experience and fewer integrations
closed firmware; scenario depends on smartphone/pairing
SafePal S1 $50 30 000+ Standalone (QR)
Android / iOS; USB
Screen + camera
battery
Binance Labs support
Affordable entry into “cold” storage
broad lists of networks/tokens
QR scenario without a cable
Outdated UI and button navigation
infrequent updates and uneven UX
compromise in build/feel

Practical check: 1) the asset exists in the official list and the support type is clear; 2) the device screen shows the address/network/amount/fee before signing.

🧩 Related topic: 2-of-3 multisig for large amounts
A 2-of-3 setup with devices from different brands reduces the risk of a “single point of failure”: device loss, mistakes, and compromise of one key.

Hardware wallet reviews: 8 models and clear trade-offs

Short reviews by scenario: universality, air-gap, BTC-only, or large screen.

How to read it: 1) “Core idea” (scenario and trust model), 2) parameters (screen/connection/software), 3) strengths and limitations, 4) final conclusion on suitability.

Trezor One — a simple “veteran” to start with

Core idea: a budget wallet without Bluetooth or battery. A common choice as a “first cold wallet,” but without a Secure Element; under physical access risks, a passphrase is usually added.

  • Connection: USB (micro-USB) • no BT/battery
  • Assets: >1,000 (BTC, ETH, LTC, ERC-20, and others)
  • Screen/control: OLED (mono) • 2 buttons
  • Compatibility: Windows / macOS / Linux; Android (OTG) • iOS — no
  • Software: Trezor Suite (Desktop + Web + Bridge)
  • Protection model: open-source; SE: no • PIN, Passphrase, offline signing

Strengths

  • Easy entry: minimal setup and a clear routine.
  • Open ecosystem: a more transparent trust model for software and updates.
  • No wireless interfaces: fewer extra attack surfaces.

Limitations

  • No iOS or Bluetooth: mobile scenarios are limited.
  • No Secure Element: under physical access, passphrase and storage discipline matter more.
  • Small screen and micro-USB: confirmations are less convenient.

Who it suits: a basic wired scenario and a low “entry threshold” into cold storage; for touch-screen or SE-class devices, other models in the lineup are more often considered.

Trezor Model T — Trezor flagship with a touchscreen

Core idea: a color touchscreen and PIN/passphrase entry on the device. Includes Shamir Backup. An “open-source without Secure Element” approach.

  • Connection: USB-C • microSD for Shamir • no BT/battery
  • Assets: ≈1,800+ • Monero is available here
  • Screen: 1.54″ color touch • on-device input
  • Compatibility: PC; Android (OTG) • iOS — no
  • Software: Trezor Suite
  • Protection model: open-source; SE: no • Shamir, CoinJoin, FIDO2

Strengths

  • Touchscreen: easier to verify address and amount, with fewer “blind” approvals.
  • Shamir Backup: convenient for splitting backup into 2-of-3 or 3-of-5 parts.
  • Open stack: a transparent software trust model.

Limitations

  • More expensive than entry-level models: the extra cost is mainly for UX and screen.
  • No iOS or Bluetooth: mainly a wired scenario.
  • No SE: if the device is lost, passphrase and storage rules matter more.

Who it suits: scenarios where touch UX and Shamir Backup matter within the most “open” trust model possible.

Ledger Nano X — a multicurrency wallet “for every day”

Core idea: a compact wallet with Bluetooth and a battery for mobile scenarios. Ledger Live covers basic operations. It has a Secure Element CC EAL5+, but the firmware is closed.

  • Connection: USB-C / Bluetooth • battery
  • Assets: ≈5,500+
  • Screen: OLED 128×64 • 2 buttons
  • Compatibility: PC; Android / iOS • limitations are possible
  • Software: Ledger Live
  • Protection model: SE CC EAL5+ • PIN, Passphrase, U2F

Strengths

  • Broad coverage of assets and integrations, including pairing with MetaMask.
  • Convenience in mobile scenarios: Bluetooth and battery.
  • Secure Element: stronger resistance under physical access to the device.

Limitations

  • Closed firmware: a different trust model than open-source.
  • Reputational disputes around the brand may matter for part of the audience.
  • On iOS, some scenarios depend on apps and platform limitations.

Who it suits: “one wallet for many things” (multinetwork + mobile scenario), if closed firmware is acceptable.

Ledger Stax — large E-Ink and a focus on confirmation comfort

Core idea: a focus on readability: curved 3.7″ E-Ink, touch controls, Qi charging, and magnets. In practice, it is close to Nano X, but screen confirmations are easier to read. SE EAL6+.

  • Connection: USB-C / Bluetooth • Qi • battery
  • Assets: ≈5,500+
  • Screen: 3.7″ E-Ink (touch) • always-on
  • Compatibility: PC; Android / iOS
  • Software: Ledger Live
  • Protection model: SE EAL6+ • PIN, Passphrase

Strengths

  • Large E-Ink: addresses and amounts are easier to read, with fewer signing mistakes.
  • Comfort: touch controls, customization, and Qi charging make daily operations easier.
  • Secure Element EAL6+: a strong hardware security base.

Limitations

  • High price: paying for UX and screen.
  • Non-removable battery: less “forever” than wired models.
  • Closed stack and extra interfaces: for a strict threat model, more minimal classes are often preferred.

Who it suits: scenarios where readability and screen confirmation comfort are the priority.

Coinkite Coldcard — BTC-only and maximum control

Core idea: a BTC wallet for “strict mode”: PSBT via microSD (air-gap), 2× Secure Element, duress PIN, and protective modes. It is often used for multisig and long-term holding.

  • Connection: USB (power) • microSD (PSBT) • no BT/battery
  • Assets: BTC-only
  • Control: OLED + keyboard • 12 keys
  • Compatibility: PC + offline via microSD
  • Software: Electrum / Sparrow / Specter
  • Protection model: 2× SE • duress PIN, Brick Me, tamper

Strengths

  • PSBT via microSD: signing without sending the transaction over a cable.
  • Protection modes: duress PIN and policies for a hard threat model.
  • Minimal interfaces: no Bluetooth and no battery.

Limitations

  • BTC only: not suitable for multinetwork setups and DeFi.
  • Higher entry threshold: the PSBT process and software pairing require practice.
  • Operation requires an external interface (Sparrow/Electrum/Specter).

Who it suits: BTC holding and multisig setups where control matters more than “everyday” convenience.

Keystone 3 Pro — QR air-gap for active DeFi

Core idea: data exchange via QR (without BT and without USB data transfer), a large 4″ touchscreen, 3× Secure Element, and protective modes. The scenario is mobile DeFi through WalletConnect and pairings with compatible apps.

  • Connection: Air-gap (QR) • no BT/USB data
  • Assets: ≈5,500+
  • Screen: 4″ touch + camera
  • Compatibility: Android / iOS • PC via QR
  • Software: Keystone Companion • WalletConnect
  • Protection model: 3× SE • Self-Destruct, Fingerprint

Strengths

  • QR air-gap: no data exchange via USB or Bluetooth.
  • Large screen: easier to verify address, amount, and network before signing.
  • QR pairings are convenient for mobile DeFi via WalletConnect.

Limitations

  • The QR scenario is usually slower than USB/BT.
  • Larger form factor than “flash-drive” devices like Nano.
  • Smaller ecosystem: experience across networks and dApp may differ.

Who it suits: QR air-gap and active DeFi operations from a phone when offline signing is the priority.

BitBox02 — minimalism and one-step microSD backup

Core idea: a compact wired wallet with file-based microSD backup, an open-source approach, and a secure chip. Multi and BTC-only versions are available.

  • Connection: USB-C • microSD (backup)
  • Assets: ≈1,500+ (Multi) • BTC-only is a separate version
  • Control: OLED • touch edges
  • Compatibility: PC; Android (USB-C) • iOS — no
  • Software: BitBoxApp
  • Protection model: SE ATECC608A • Anti-Klepto, U2F

Strengths

  • microSD backup: fast backup and recovery without a manual routine.
  • Balance: open-source approach plus a secure chip.
  • The BTC-only version simplifies the feature set and risk model.

Limitations

  • No iOS or Bluetooth: a wired scenario with PC/Android.
  • Lower network support than Ledger: the asset list must be checked for relevance.
  • The touch edges require adjustment, especially after “buttons.”

Who it suits: careful storage with simple backup and minimalist UX as the priority.

NGRAVE ZERO — a “desktop vault” with EAL7 and QR mode

Core idea: an offline device in a “mini-smartphone” format: QR exchange, large screen, EAL7 SE, tamper sensors, and a physical Graphene backup. A niche class focused on physical risks.

  • Connection: Air-gap (QR) • USB-C: charging/firmware
  • Assets: ~1,000+
  • Screen: 4″ touch + camera • 480×800
  • Compatibility: Android / iOS
  • Software: Liquid • WalletConnect; QR integrations
  • Protection model: SE EAL7 • tamper sensors, NGRAVE OS

Strengths

  • Strong hardware focus: EAL7 and tamper control for physical threats.
  • QR air-gap: no wireless channels and no USB data exchange.
  • Large screen: easier offline approval and lower error risk.

Limitations

  • Very high price: a niche device class.
  • Less mainstream product: fewer familiar guides and less field-tested experience.
  • Asset coverage is usually lower than that of leading multicurrency devices.

Who it suits: a strict threat model and large amounts, where offline signing and physical device resilience are the priority.

Niche models and worthy alternatives

These are not the “main hits,” but devices for specific scenarios: ultra-budget, Trezor with SE, BTC-only with better UX, open-source for a technical approach, NFC card for everyday use.

SafePal S1

Format: air-gap via QR (USB is optional, depending on the scenario).

Assets: more than 30,000 tokens/networks are claimed — actual support depends on the SafePal App.

Strong side: low entry price into cold storage + broad network coverage on the lists.

Limitation: simpler UX and closed firmware — the trust model differs from open-source devices.

Who it suits: scenarios where minimum price and broad network lists matter, and the QR process is acceptable in terms of speed.

Trezor Safe 3

Format: wired (USB), without Bluetooth.

Ecosystem: Trezor Suite (PC + Android via OTG).

Strong side: the “Trezor experience” + Secure Element in a more affordable lineup.

Limitation: no iOS/BT and no touch controls — confirmation and navigation are less “mobile.”

Who it suits: scenarios where the Trezor interface and an SE chip for physical access risks matter.

Blockstream Jade

Format: hybrid modes — USB / Bluetooth / QR.

Focus: an open-source approach and BTC scenarios (often through Green and compatible wallets).

Strong side: a choice of connection mode for the situation (USB for simplicity, BT for mobility, QR for isolation).

Limitation: the absence of a “classic” Secure Element — for some scenarios, this is decisive.

Who it suits: an open-source approach and flexible connection modes without being tied to one scenario.

Foundation Passport

Format: BTC-only with QR air-gap and a focus on clear UX.

Integrations: most often paired with Sparrow/Specter and other BTC tools.

Strong side: a strict BTC focus without a “purely technical” interface.

Limitation: high price and strictly BTC-only (without an “alts just in case” scenario).

Who it suits: BTC-only storage with QR air-gap when readability and navigation are the priority.

Tangem Wallet

Format: an NFC “smart card” without a screen — managed through a smartphone.

Security: Secure Element (often listed as EAL6+); firmware is closed/unchangeable.

Strong side: minimal entry threshold and a fast everyday scenario.

Limitation: without a screen on the device, there is no hardware-level verification of address/amount; dependence on the smartphone increases.

Who it suits: scenarios where speed and card form factor matter, while autonomy and screen verification are secondary.

Selection guide: a hardware wallet for profile and scenario

The choice comes down to three parameters: assets (BTC-only or multinetwork), signing (USB / Bluetooth / QR), and risk scale (daily operations or large amounts).

Short selection logic:

  1. Phone: Bluetooth (Ledger) or QR air-gap (Keystone / SafePal / NGRAVE).
  2. Fewer “blind” approvals: a readable address and amount on the screen (a larger screen or a more transparent interface).
  3. Large amounts: priority goes to multisig + distributed backups schemes, then to model selection.

1) For beginners and everyday use

  • Trezor Safe 3 / Trezor One — a wired scenario and the Suite ecosystem; protection against substitution depends on checking the address on the screen.
  • Ledger Nano S Plus — budget-friendly and broad coin support in a USB scenario.
  • SafePal S1 — many networks for minimal cost with QR signing; slower than USB/BT.

2) DeFi/NFT and active use

  • Ledger Nano X / Ledger Stax — Bluetooth, Ledger Live, and integrations for multinetwork use.
  • Keystone 3 ProQR air-gap for mobile DeFi through WalletConnect and compatible apps; QR exchange adds steps and reduces the number of data exchange channels.
  • BitBox02 Multi — microSD backup and ecosystem; network coverage should be matched against target tokens.

3) Large amounts and a strict threat model

  • 2-of-3 multisig — lower risk of a “single point of failure”: different brands + distributed storage of keys and backups.
  • NGRAVE ZERO — QR air-gap and a focus on protection (more expensive and less “everyday” in format).
  • Routine matters more than the model: passphrase, metal backup, recovery test, and checking the address/amount on the screen.

Critical during every signature: check the address, network, and amount on the device screen. The seed is stored offline (preferably in metal), and PIN/seed/passphrase are never shared with third parties.

🧯 Related DeFi topic: approval/allowance and revoking permissions
In DeFi, a common cause of loss is granted permissions (allowance). The material covers spender, limits, and revoking access (revoke).

Questions and answers (FAQ)

Seed phrase backup for a hardware wallet: metal, passphrase, or Shamir — what to choose?

The basic level is metal and a second set in another place. For large amounts, passphrase is added or Shamir 2-of-3 is used; a separate step is a one-time recovery check on a “clean” device. A detailed breakdown of backup strategies is covered in a separate article: seed phrase: storage and backups.

An infected PC/smartphone and a “clean” hardware wallet: where does the risk remain?

Private keys do not leave the device, but an infected host can replace the address and details in the interface. The practical response is to stop operations and move to a clean environment; in case of doubt, a new seed + passphrase scenario and transfer of funds is used.

DeFi/NFT through a hardware wallet: what matters on the screen during signing?

It is critical that the device screen clearly shows the operation details and action type (for example, transfer or approve), and in EVM — who the spender is and what allowance is being granted. A deeper look at approvals and revoking permissions is covered separately: approval/allowance: checking and revoke.

What should be checked on the device screen before signing?

To whom (address), what (asset/amount), where (network/chainId), how much (fee). In EVM, separately: transfer or approve, and who the spender is. If the operation is unclear, signing is postponed.

When does multisig with hardware wallets make sense?

For large amounts, a 2-of-3 setup on devices from different brands reduces the risk of a single point of failure. The scenario conditions are separate storage of keys and backups, and a test of losing one device. A step-by-step setup and cross-vendor nuances are covered separately: 2-of-3 multisig: principles and setup.

Inheritance and a hardware wallet: what should be planned?

There are two layers: legal documents and a separate “technical instruction” (where the backups are and what the sequence of actions is). The seed phrase is not placed in a will; the scenario must work without the owner’s participation.

Final: cold storage that actually works

A hardware wallet stores private keys inside the device and signs transactions offline. Screen confirmation reduces the risk of address replacement and “signing the wrong thing.”

A practical formula: keys offline + backup stored separately + checking on the screen — the basic set that matters more than a “top model.”

The seed phrase must not appear in digital form (photos/scans/cloud/input on a PC). If the seed has been “digitized,” the risk of compromise is considered elevated.

Recommended

  • Official seller and self-initialization of the device.
  • The seed is kept offline only (preferably in metal), and copies are stored separately.
  • PIN always; for large amounts, passphrase is added and stored separately.
  • Before signing, check the address/network/amount/fee (in EVM — spender for approve).

Not recommended

  • Second-hand purchases, pre-configured devices, and non-obvious discounts.
  • Photos/screenshots of the seed, sharing it, and cloud storage.
  • Keeping the device and backup together as one point of failure.
  • “Blind” signing and unlimited allowance when it is unclear who is receiving the rights.

Mini launch algorithm

  1. Purchase: official seller → check the packaging and contents.
  2. Initialization: the seed is created on the device → written down offline → second set.
  3. Access: PIN → if needed, passphrase (stored separately).
  4. Check: a small test transfer there and back.
  5. Separation: a separate address for storage and a separate “operational” one for DeFi.
  6. Maintenance: updates only from official sources + periodic review of permissions (allowance).

Final thought: security comes not from the “most expensive wallet,” but from routine: offline backup, separate storage, and checking on the screen.

Found this article useful?

Subscribe to our updates to not miss new reviews and ratings

View All Exchanges →