Trezor review: the original open-source hardware wallet, a decade on

Czech company SatoshiLabs launched the Trezor One in 2014, a device widely regarded as the first dedicated hardware wallet sold commercially. Its present range consists of three levels: Trezor Safe 3 at $59 with a monochrome display and physical buttons; Trezor Safe 5 at $129 with a 1.54" colour touchscreen, Gorilla Glass and a MicroSD backup slot; and Trezor Safe 7 at $249 with a 2.5" colour touchscreen, Bluetooth 5.0+, Qi2 wireless charging, an aluminium body and IP54 protection. Each model uses USB-C and works with the free Trezor Suite on desktop, Android or the web.

The defining feature of Trezor remains its open-source approach: both device firmware and Trezor Suite source code are publicly available on GitHub for independent review. The hardware, however, has evolved. Every model in the current range now contains a certified EAL6+ Secure Element, a certification level also found in EMV payment cards and government ID chips. That change directly addresses earlier criticism of older Trezor devices, which depended on a general-purpose microcontroller rather than a dedicated secure element.

Every model in the current Safe 3/5/7 lineup includes a certified EAL6+ Secure Element, storing private keys in tamper-resistant hardware isolated from the general-purpose chip that runs the open-source firmware. The flagship Safe 7 goes further with SatoshiLabs' own TROPIC01 secure chip alongside the certified element — a design the company describes as built specifically to be independently auditable, in keeping with Trezor's open-source philosophy.

Recovery uses a standard 12- or 24-word BIP39 seed phrase, with optional passphrase protection for hidden wallets, and Trezor additionally supports Shamir Backup (SLIP39), which splits a recovery seed into multiple independent shares so no single share alone can reconstruct your funds. Because both firmware and app are open source, security researchers and competing wallet makers alike have been able to audit Trezor's code for over ten years — one of the longest continuous public-scrutiny track records in the industry.

The device's security history includes a well-documented 2020 finding: researchers at a hardware security lab demonstrated a voltage-glitching attack that could extract the seed from earlier Trezor One/Model T hardware given extended physical possession and lab equipment — not a remote exploit. Trezor's mitigation is a strong passphrase, which keeps funds safe even if a device is physically compromised. Separately, in August 2026 Trezor disclosed that ShipMonk, its third-party order-fulfilment provider, suffered a breach exposing 80,689 customers' shipping details; Trezor has stated that no customer funds, devices or wallet software were compromised, but warned affected users to expect a wave of phishing attempts using the leaked contact data.

Trezor covers three price levels, from an affordable entry model to a premium wireless flagship, while maintaining the same core secure-element certification throughout the range.

Trezor Suite (desktop app, web app and Android companion) supports thousands of coins and tokens natively, with Bitcoin, Ethereum and other EVM chains as first-class citizens; broader multi-chain coverage is available by connecting Trezor as a signing device to third-party wallets such as MetaMask, Electrum or MyEtherWallet, a pattern common across the industry for wallets that don't try to natively support every chain themselves.

Built-in features include a portfolio dashboard, an in-app coin exchange/swap via third-party integrated providers, and — distinct from most competitors — a native Tor connection option inside Trezor Suite for users who want to route their wallet traffic through the Tor network for additional network-level privacy.

Vs. Ledger: Ledger's secure-element firmware remains closed source, a long-running community criticism, while Trezor has published both firmware and app source since day one. Ledger has greater brand scale and a larger asset/app catalogue, and had its own much larger 2020 customer-database breach (over a million emails, hundreds of thousands of physical addresses) — a useful point of comparison for Trezor's smaller-scale August 2026 incident.

Vs. OneKey (also reviewed on this site): OneKey's pitch has been "open source like Trezor, secure-element-hardened like Ledger" — a direct response to Trezor's older models shipping without a certified secure element. Current-generation Trezor hardware (Safe 3/5/7) now ships with EAL6+ certification across the board, closing that gap, while retaining a decade-longer public track record and a lower entry price ($59 vs. OneKey's $79 Classic 1S Pure).

The August 2026 data breach is recent and real. Even though Trezor is correct that no device or wallet software was touched, 80,689 customers now have their name, address and contact details in the hands of whoever obtained the ShipMonk data — a meaningful phishing and even physical-targeting risk that deserves to be taken seriously, not waved away.

"Open source" took a decade to include a certified secure element. For years, Trezor's answer to "why no secure element?" was that open-source firmware was itself a sufficient defense. Competitors and some security researchers disagreed, and Trezor's current hardware effectively concedes the point by adding one.

Physical attacks are not purely theoretical. The 2020 voltage-glitching research was conducted by a professional lab, not a random attacker, but it proved current-at-the-time hardware could be compromised given sustained physical access — a reminder that "open source" and "physically unbreakable" are not the same claim.

iOS support depends on the model. Safe 7 supports full Trezor Suite use on iOS through Bluetooth, while Safe 3, Safe 5 and older models have more limited iPhone support and generally require desktop or Android for signing transactions.

This isn't Trezor's first phishing-adjacent incident. An earlier newsletter-related data exposure around 2020–2021 was also followed by a wave of phishing attempts against Trezor users — a pattern worth knowing about if you're a long-time customer receiving unexpected "security" emails.

Separate September 2026 email-provider incident. On September 10, Trezor disclosed that a breach at Brevo affected its opt-in newsletter database of roughly 347,000 email addresses and was used to send a phishing email from Trezor's account. Trezor says no other Trezor system was touched.