Timechain
Search Donate
Sponsors
Cake Wallet, private Bitcoin and Monero walletSideShift.ai, no-signup crypto exchangeOrangeRock, non-custodial bitcoin spot and perps tradingBraiins Hashpower, buy and sell Bitcoin hashrateLayer Two Labs, drivechains for Bitcoin
S15 · E69 · 2:13:38

S15 E69: Zach Herbert on Passport Prime by Foundation Devices

17 Dec 2024 · by Vlad Costea · 15 min read
Published at block 875,086
Cover art for S15 E69, Zach Herbert on Passport Prime by Foundation Devices
MP3 (248 MB)

The audio keeps playing while you browse the site.

Read the transcript20,930 words, every line links to the minute it was said

Ask this episode

Embed this episode

Paste this on your site or blog. It is a small player that streams from here: no trackers, no cookies, no ads.

Preview the player

I had Zach Herbert back on the show because he finally had a third-generation device to talk about, and because I wanted him to talk me out of my Trezor Model T. The last time we recorded at length was the summer of 2021, when Foundation was still new and he sent me a Founders Edition Passport with a copy of the Bitcoin white paper. In Season 8 Episode 5 that first device still felt like a bet. By December 2024, Ledger and Trezor were still the default names people recite, and Zach was no longer trying to win the old hardware-wallet race on the old terms.

"We're announcing our third-generation Passport, and we're specifically not calling it a hardware wallet. We're calling it a personal security platform."

The product is Passport Prime. If the pitch works, it is not only a place to keep a seed. It is also a second-factor device, an encrypted drive, and later an app platform. I told him the Steve Jobs line almost immediately: today we are introducing a device that holds your keys, a device that stores your files, and a device that does your 2FA codes, and they are all the same object. He said they had thought about using that joke in the keynote and then chickened out, because they were not sure enough people would get the reference.

Not a Hardware Wallet

Passport Prime is a compact touchscreen device with a 3.5-inch display, closer to a credit card in footprint than to a phone. The first firmware, which they were aiming to ship at the end of Q1 2025, was not the full app store dream. It was already a bundle.

"What we're shipping in end of Q1 of 2025 is the same Bitcoin wallet functionality that we have on Passport today. We're shipping six-digit 2FA codes and FIDO security keys, basically YubiKey functionality, via NFC and USB. We're also shipping 50 gigabytes of encrypted storage... and we're also shipping a seed vault."

The more ambitious half of the pitch comes later. Zach said they would have to build a developer ecosystem from scratch, and that the broader platform would arrive mid-2025. Until then, the honest description is this: it is a much more computer-like device than a traditional hardware wallet, running a real operating system, with a plan to let other people add features over time.

That is why they do not want the old category name. A hardware wallet, in the way most people use the phrase, is a small vault with a narrow menu. Passport Prime is trying to be the vault, the YubiKey, the encrypted USB stick, and eventually the place where other security tools live. Whether that is a platform or just a very full backpack is the question the rest of the interview tries to answer.

KeyOS, Xous, and a Microkernel in Rust

The software story is the part I found most interesting, because it is not a reskin of what Ledger or Trezor already ship. Passport Prime runs an operating system Foundation calls KeyOS. It is built on Xous, a microkernel written in Rust that came out of Bunnie Huang's Precursor project, the hardcore open-source secure communications device that uses an FPGA to simulate a RISC-V core.

Zach's version of the genealogy is refreshingly unembarrassed. They did not invent a microkernel in a vacuum. They found a serious hobbyist project, became some of its largest GitHub contributors, ported it from RISC-V to ARM because the security-grade parts they can actually buy are ARM, added color graphics, and then built the apps. He described KeyOS as a distribution of Xous, the way Ubuntu is a distribution of Linux.

Microkernels matter here because the whole point is isolation and resilience rather than one big blob of permissions. He mentioned Google's Fuchsia and the desktop work around Redox as proof that this idea has been circling serious engineers for years, and then made the claim that it has not really shown up in a consumer device before. I cannot audit that historical claim from a podcast chair. I can say that "we started from an open-source project and then did the unglamorous port" is a better story than "we invented a secure element and you will have to trust the brochure."

He also knew the hate mail would arrive anyway. Someone would say they copied NVK, or that they followed a trend, or that RISC-V was in the water. His answer was that they had been working on this product for about two years, from a plan they already had, and that every time a competitor shipped something during that wait he was relieved, because it was not this.

Stalled Innovation, and a 2.5 Percent Ceiling

Zach's tour of the current market is unkind, and I think it is mostly fair. Coldcard, after years of saying cameras were a bad idea, added QR support in a bulky way with a scanner module. Trezor, the company he said he respects most in the space because it stayed with the open-source mission, shipped Safe 3 and Safe 5 devices that feel to him like modernized reskins: same form factor, same software, a more current look, and a secure element on the Five that never compelled him to upgrade. Ledger's Stax and Flex, in his telling, took the internals of the Nano line, including Bluetooth, and put them behind a bigger e-ink screen, then kept the same app-loading ritual and the same Ledger Live gravity well.

"I think that, in summary, innovation seems to have stalled in the hardware wallet space."

The adoption number he used is the one that should make the whole industry sit down. Foundation's estimate was that only about 2.5% of crypto users actually use a hardware wallet. Maybe 10 to 15 million devices have been sold in a decade. Ledger, when it launched Stax, was talking about something like 7 million devices over that whole period, and had not cracked a million additional units in the year and a half after the announcement. Trezor has said "over a million." Everyone else, including Coldcard, BitBox, Jade, and Foundation, is fighting over scraps.

I still use a Trezor Model T, and I use it in a slightly feral way: I wipe it and recover from seed so I am not storing anything in flash. I like the touchscreen because it is faster to type, like an old Nokia. Zach's challenge in this episode was to convince me to leave that ritual behind. He did not do it by insulting Trezor. He did it by arguing that the entire category may be the iPod: beloved by enthusiasts, irrelevant to the next 100 million people, and due to be absorbed into something that does more than one job.

Still Bitcoin-Centric, Now a Platform

Then I roasted him, because the press release mentioned Cake Wallet and Monero, and because I can play the intolerant Bitcoin-only character when a guest walks into it.

"Foundation is still, and has always been, as I describe it, a Bitcoin-centric company."

The distinction he wants is between the company and the platform. Envoy, the companion app, remains a simplified Bitcoin wallet. Passport, the existing air-gapped Bitcoin-only device, would keep being sold. Passport Prime is supposed to be an open platform, "the polar opposite" of Ledger's closed one, where a third party can ship an app to its own users without waiting for Foundation's permission, and later also list it in an Envoy catalog. SSH signing, PGP signing, and other non-coin security tools are part of the fantasy. So is Monero, if Cake Wallet can do on Prime what it has not been able to do cleanly on Ledger.

I put the hater version of the argument to him directly: maybe they could no longer compete with Coldcard at being Bitcoin-only, so they decided to become a platform for every other coin instead. His answer was more interesting than a denial. He thinks the entire hardware-wallet industry is losing, including the purest products in it, because the adoption is terrible. A device that can only ever be a Bitcoin vault, marketed through drama and annual revisions, is not how you get millions of people into self-custody. He would rather leapfrog Ledger and Yubico by becoming infrastructure other companies build on.

You can disagree with that and still hear the empirical point. If the best Bitcoin-only wallets are stuck in a 2.5% ghetto, "stay pure and lose" is also a strategy. It is just a smaller one.

Two Hundred and Ninety-Nine Dollars, Plus the Things It Replaces

"Passport Prime is $299." Passport Gen 2 stays at $199, which is already, Zach said, about as low as they can go while manufacturing in the United States and keeping a decent margin. The Founders Edition had been $299, Gen 2 launched at $249, and they brought it down as production scaled. He did not want to cut the Gen 2 price just because a new device exists, and he hoped not to raise the Prime price either, maybe adding a higher-storage tier someday if people actually want to use it as a little hard drive.

The $299 comparison he wants you to make is not only with Ledger Flex at $249 or Stax at $399. It is with the pile of objects Prime is trying to replace. "If you start to compare the cost of the different devices that Passport Prime is replacing, I think it looks like a pretty good deal." A hardware wallet, two YubiKeys (because you cannot back one up, and some services, including Apple's Advanced Data Protection, want two), and a hardware-encrypted flash drive add up, in his math, to something like $400 to $600 before anyone has written a third-party app.

The first 1,000 Early Access units came with a lifetime subscription to Envoy Plus. Everyone else who bought a Prime would get a six-month trial. General-availability buyers could reserve on the site and wait for the non-beta firmware. The early batch was aimed at March 2025, starting on beta firmware and then moving to a v1 release.

I did not hear this as cheap. I heard it as a bundled price that only makes sense if you actually use the extra jobs. If all you want is a stateless signer, $299 is a lot of computer.

Shamir Cards, Magic Backups, and a Subscription That Tries Not to KYC You

The backup story is where Prime stops looking like a thicker Passport and starts looking like a phone.

"By default we're doing Shamir, which is pretty cool. I know Trezor has that as an option... What we're doing is, if you opt into it, we're doing a 2-of-3 Shamir, and two of the shares go on the included cards." Those NFC key cards are already in the box, in Faraday sleeves, written with the device's own NFC antenna under the glass on the back. The third share can live in the Envoy app. If you lose the device, you tap a card and let the phone supply the second share. Foundation also ships a third physical card, so privacy-minded users can keep all three shares in their own hands and never use the phone or the server at all. You can also plug in a flash drive or SD adapter and take manual encrypted backups, because the device can read and write external storage.

The optional subscription, Envoy Plus, is what they call Magic Backups. It was planned for mid or late 2025. "What our subscription service allows you to do with Magic Backups, as we call it, is the device takes an encrypted backup of all of the metadata and settings, so not the seed words." That backup is encrypted with the seed and sent through Envoy to Foundation's servers. "The subscription is essentially paying us, it'll be five bucks a month, to store this backup for you." The user identifier, in Zach's description, is a hash of the seed. "The hash of your seed is actually the user ID, which is really cool, and we don't have any other data about you." If you use Tor in Envoy, they are not trying to attach a name to that blob. Restore is supposed to feel like iCloud without handing Foundation a copy of the keys: tap a card, download the encrypted blob, decrypt it on device, and get your 2FA codes, security keys, seeds, and app data back.

I asked the obvious question: a subscription usually means KYC, credit cards, email addresses, and a support desk that knows who you are. Zach said they would know that someone paid, because that is how a paid service works, but they were still designing a way to redeem a code so that the email and payment method would not be joinable to the seed-hash and the backup blob. They delayed the paid launch, in part, because they had not finished the private-payment side.

I am willing to be impressed by the intention and still wait for the implementation. Ledger Recover became a cautionary tale by putting seed material in the wrong conversation. Zach kept repeating that Magic Backups are not that: the words stay off the server. If they ship what he described, it is one of the more serious attempts I have heard from a hardware company to offer convenience without turning the backup into a customer dossier. If they fail, it will fail in the usual way, with a database and an invoice.

Built in the United States, Bluetooth in a Tunnel

Foundation still manufactures in the United States. I asked why, when the margins would be better somewhere else. "It's purely selfish," Zach said. He would not buy a device like this if it were made in China, or in Asia, or specifically provisioned by Foxconn. He wants to know who built the board.

That led to a useful correction on Ledger. I had seen chatter about Hungary and China. Zach had a Flex box in front of him: designed in France, made in Vietnam, by Foxconn. The newer secure touchscreen Ledgers, Stax and Flex, are no longer made in France or even in Europe. He thinks Tony Fadell, brought in after the iPod and Nest chapters of his career, pushed them toward that kind of scale, and that the curved e-ink panel then punished them with yields and delays. Flex, at $249, is the stock-panel rescue device. Stax, at $399, is the expensive custom one.

The other polarizing choice on Prime is Bluetooth. Zach admitted the stack has closed-source blobs and that this bothers security people, himself included. Their mitigation is architectural. The Bluetooth radio is a separate chip, not baked into the processor that runs KeyOS, unlike the ESP32-style designs that scare him. The first pairing is still air-gapped: you scan a QR code and exchange encryption keys with the phone before anything important moves. After that, the radio only ferries already-encrypted, signed blobs. They call the protocol Quantum Link, and Zach said it uses post-quantum encryption. His line was, "it gives the security of an air gap while having the convenience of Bluetooth."

The rest of the hardware, as he described it, is meant to be auditable. KeyOS is open source, under GPL, like the current Passport. The chassis and a fairly complex board will be released under an Open Hardware Licence. The main chip is a Microchip security MPU with ARM TrustZone, running KeyOS at about 500 MHz, with active tamper switches that wipe the device if you open it. There is still a Microchip 608-series "dumb" secure element for key slots, the same family used on current Passport, Coldcard, and BitBox, with no application code executing on it. The only closed-source residue he would not talk his way out of is down in that Bluetooth stack.

He would not put a decade-long support guarantee on Passport 1 and 2 in a podcast, but he did say years, maybe a decade, and that they had already back-ported a color-era UI to the monochrome Founders Edition because they did not want to strand early customers. Gen 2, he said, would probably still be sold at least through the end of 2025. They will watch how sales split between the Bitcoin-only wallet and the platform, and maybe do an air-gapped refresher if that market still wants one.

Pre-Selling a Thousand, and a Homebrew Future

At the end I asked the only question that mattered to me: did he sell me, or am I keeping the stateless Trezor? I told him the truth. I still like the Trezor. I am curious to try the bigger screen, especially if recovering a seed on it is less painful, and I like that Prime can also do temporary or ephemeral seeds, the SeedSigner-style workflow, for people who want to stay stateless on a fancier machine.

"We are pre-selling a thousand of these as part of our Early Access program, which we're aiming to ship in March." Those first units would start on beta firmware and then move to a v1 release. "Our first launch partner is Cake," which means Monero support is not a hypothetical appendix. The mid-2025 SDK is the real test. I told him about a Nintendo 3DS I had lying around, dead as a product, suddenly interesting again once unofficial firmware and Homebrew software landed on it. I want Prime to attract that kind of builder, not only a list of coins. Zach was happy to start with Doom, or at least with haptic feedback standing in for a speaker. A 500 MHz security computer that can only run the apps Foundation imagined on day one would just be a very expensive wallet.

Looking back from 2026, I later sat down with Zach again in Season 17 Episode 24, after some of this had met actual users. I am not going to retrofit that later conversation into this one. In December 2024, Passport Prime was still a pre-order and a thesis: that hardware wallets, as a category, had stalled, and that the way out was a device people would keep in their pocket because it did more than sign a PSBT.

Listen to Zach Herbert on Spotify, Apple Podcasts, YouTube, and More

Time stamps:

  1. 00:01:49– Announcing the New Passport Prime
  2. 00:03:17– Competitive Landscape of Hardware Wallets
  3. 00:03:17– The Challenge to Replace Trezor
  4. 00:05:25– Introduction of the Passport Prime
  5. 00:06:54– Features of the Passport Prime
  6. 00:08:07– Expanding Digital Security
  7. 00:08:34– Comparison to Steve Jobs' Presentation
  8. 00:10:34– Open Source Platform Development
  9. 00:11:46– Microkernel Operating System
  10. 00:14:31– Unique Approach to Hardware Wallets
  11. 00:15:20– Development Timeline for Passport Prime
  12. 00:17:17– Competitor Analysis
  13. 00:21:03– Innovation (and Stagnation) in Hardware Wallets
  14. 00:24:33– Foundation's Bitcoin-Centric Approach?
  15. 00:25:57– Comparison with Ledger's Closed Platform
  16. 00:28:16– Developer Ecosystem and App Distribution
  17. 00:29:50– Hardware Wallet Industry Challenges
  18. 00:30:36– Market Position of Ledger & Trezor
  19. 00:35:26– Foundation's Still Manufactured in the USA
  20. 00:35:41– Manufacturing Changes in Ledger
  21. 00:38:23– Pricing Strategies of Hardware Wallets
  22. 00:40:09– Passport Prime Pricing and Features
  23. 00:42:26– Competing with YubiKey
  24. 00:43:05– Envoy Plus Subscription Service
  25. 00:44:40– Passport Prime Packaging
  26. 00:44:54– Shamir Backup Method
  27. 00:46:02– Magic Backups Feature
  28. 00:48:11– Privacy Concerns with Subscription
  29. 00:51:20– Backup Management Options
  30. 00:56:14– Open Source Elements
  31. 01:00:19– Bluetooth Security Measures
  32. 01:03:33– Support for Previous Generations
  33. 01:03:55– User Interface Adaptation
  34. 01:04:55– Product Support Commitment
  35. 01:05:30– Price Stability for Passport Devices
  36. 01:07:23– Future Mobile Phone Plans
  37. 01:09:31– Operating System Development Challenges
  38. 01:11:53– Optimization in Software Development
  39. 01:13:57– Microkernel Operating System Vision
  40. 01:17:02– Integration of Open Source Components
  41. 01:18:28– Concerns Over Tropic Square's Project
  42. 01:22:37– Future of Foundation Devices
  43. 01:23:27– Open Source Benefits
  44. 01:24:11– Firmware Updates and Security
  45. 01:25:18– Backup Solutions
  46. 01:26:55– Critique of Competitors
  47. 01:29:23– CoinJoin Feature Discussion
  48. 01:31:28– Regulatory Impact on Development
  49. 01:32:39– Integration with Full Nodes
  50. 01:34:03– Challenges of Node Development
  51. 01:36:15– Ambitious Hardware Ideas
  52. 01:37:05– AI Integration with Home Servers
  53. 01:39:45– Listing and Integration Policies
  54. 01:43:03– The Hardware Wallet Ecosystem
  55. 01:44:12– Sandboxing Applications
  56. 01:45:51– Customization and User Experience
  57. 01:46:25– Gaming Features in Hardware Wallets
  58. 01:49:29– Libertarian Principles in Bitcoin
  59. 01:50:34– Market Competition and Ledger
  60. 01:51:44– Review Unit for Mandrik
  61. 01:52:27– Durability of Passport Prime
  62. 01:54:34– Conspiracy Theories About Foundation
  63. 01:56:53– Principles of Decentralization
  64. 01:58:06– Nuance in Cryptocurrency Criticism
  65. 01:59:54– Software and Distribution Analogies
  66. 02:01:12– Shifts in Hardware Wallet Design
  67. 02:03:00– Introduction to First Principles
  68. 02:03:35– Privacy Magazine Mention
  69. 02:05:00– Discussion on Limited Editions
  70. 02:05:55– Existential Questions in Bitcoin
  71. 02:07:11– Critique on Bitcoin Upgrades
  72. 02:08:10– Jaded Perspectives on Bitcoin
  73. 02:10:48– Pre-Selling Passport Prime
  74. 02:12:09– Customization and Future Updates
  75. 02:12:45– Homebrew Support
  76. 02:13:28– Closing Remarks

Verify this episode

This is the SHA-256 fingerprint of the MP3 as published here, timestamped in the Bitcoin blockchain with OpenTimestamps. If one byte of the file changes, the fingerprint changes.

SHA-256
fdc842ed4db59082a8d1bf095ee455597f69573395ec92992d4da0f54b46be28
Proof
2024-12-17_s15_e69_zach_herbert_on_passport_prime_the_future_of_hardware_wallets.mp3.ots · all episode hashes (proof)
Status
Anchored in Bitcoin block 974,312
Timestamped
After release. The episode came out on 17 Dec 2024, but this proof was made on 3 Nov 2026 and anchored in block 974,312, so it proves this exact file existed by 3 Nov 2026, not on the release date.
How to check it yourself
  1. Download the MP3 and the proof into the same folder.
  2. Run sha256sum 2024-12-17_s15_e69_zach_herbert_on_passport_prime_the_future_of_hardware_wallets.mp3 (on a Mac: shasum -a 256 2024-12-17_s15_e69_zach_herbert_on_passport_prime_the_future_of_hardware_wallets.mp3). The result must match the SHA-256 above.
  3. Install the OpenTimestamps client (pip install opentimestamps-client) and run ots verify 2024-12-17_s15_e69_zach_herbert_on_passport_prime_the_future_of_hardware_wallets.mp3.ots. It hashes the MP3 again and checks the proof against your own Bitcoin node, then prints the block and the date the file existed by. Without a node, ots info 2024-12-17_s15_e69_zach_herbert_on_passport_prime_the_future_of_hardware_wallets.mp3.ots shows the block height to look up in any block explorer.

Share: X · Email · Link

Related

Comments

Comments run on my own server. No trackers, no tracking cookies.

0:000:00