Thunderbird is the closest in spirit. Epistles is the next protocol generation.

Thunderbird and Epistles share more than they disagree on: cross-platform, multi-account, privacy-respecting, made by people who treat the inbox as a tool rather than a feed. The honest comparison is not about which project is better; it is about which set of trade-offs fits the mailbox you actually have.

What Thunderbird does well

Thunderbird has been a working email client since 2003. It is open source, governed by MZLA Technologies as a Mozilla Foundation subsidiary, and supported by donations rather than subscriptions. Those facts are the reason Thunderbird outlasted industry shifts that buried better-funded clients, and the reason a careful user can read its source and decide for themselves what the program does with their mail.

The desktop story is genuinely complete. Windows, macOS, and Linux ship from the same codebase, with installers, package repositories, and Flatpaks for the Linux distributions that prefer one over the other. Multi-account, unified inbox, tag systems, message filters, and a built-in calendar that grew out of the old Lightning extension all work without subscription gates. PGP, once an Enigmail add-on, is now in the box.

The mobile story is recent but real. Thunderbird for Android, the project formerly known as K-9 Mail, became the official Android client in October 2024 and is on version 8 as of this writing. iOS is in active development, with a TestFlight build for early testers and a public roadmap targeting a first release later in 2026. Thunderbird is not pretending iOS is shipping; they are building it in the open.

The add-on ecosystem, two decades deep, is its own argument. If your workflow depends on a specific extension, no other open-source mail client matches Thunderbird's catalogue.

Where Thunderbird stops, and where Epistles continues

Thunderbird connects to most providers over IMAP and SMTP. It is a deliberate portability choice: IMAP works everywhere, and the project has always optimised for not being captured by any one vendor. The cost is that provider-specific concepts arrive flattened. Gmail labels become folders pretending to be labels, server-side filters and categories are read-only at best, and JMAP support is on the 2026 roadmap rather than shipping today.

Epistles took the opposite bet. Gmail through the Gmail API, Microsoft 365 through MS Graph, Fastmail through JMAP, ProtonMail through Proton's API with OpenPGP decryption on device, IMAP and SMTP for everything else. We pay the price of writing five adapters instead of one. In return, labels stay labels, threads stay threads, and push arrives in each provider's native protocol: Gmail Pub/Sub, MS Graph subscriptions, JMAP push.

iOS is the other live difference. Thunderbird for iOS is being built; Epistles ships on iOS today, from the same TypeScript core that runs the Mac and Linux builds. If your inbox lives on a phone in your pocket, that gap matters this year, and it will close on Thunderbird's side later.

And there is the Cloud Vault. Thunderbird is per-device by design: each install configures each account locally. Epistles wraps every connected-account credential with a key derived from your Epistles password and stores opaque ciphertext bytes on our servers we cannot read. Sign in on a second device, type your password once, and the same accounts are already there. Proton is carved out: its mailbox passphrase and OpenPGP material live in the OS keychain only.

Where Thunderbird wins

Open source today. This is the largest single difference, and we will not minimise it. Thunderbird's source is readable, auditable, forkable, and has been for twenty years. Epistles is in the middle of migrating its working tree to a public repository at github.com/epistlesapp/epistles. That migration is real work in progress, not a gesture, but as of today the source is not yet open. If open-source-now is a hard requirement, Thunderbird is the right answer.

Free. Thunderbird costs nothing. Epistles is free for three connected accounts and $35 a year past that. Both are reasonable models for what they are; only one is donation-funded.

Calendar as a deep first-class feature. Epistles reads CalDAV, JMAP, and Google calendars alongside the mail. Thunderbird's calendar is its own product, with two decades of accumulated capability around recurring events, multiple calendars, invitation handling, and add-on extensions. If calendar depth is the wedge, Thunderbird is ahead.

Add-ons. A mature ecosystem of extensions, themes, and filters. Epistles has none of this surface area today.

Mozilla Foundation governance. A non-profit steward with a long track record on user-rights questions. We respect that standard and intend to live up to it; Thunderbird already has.

The matrix, plainly

Platforms shipping today

Thunderbird: Windows, macOS, Linux desktop; Android. Epistles: macOS, Linux, iOS, Android. (Thunderbird iOS in active development, target later 2026. Epistles Windows and web in development.)

Email providers and protocols

Thunderbird: IMAP and SMTP for almost all providers, with OAuth where supported; JMAP on the 2026 roadmap. Epistles: Gmail API, MS Graph, JMAP, Proton, IMAP/SMTP, each as a native adapter.

Cross-device account sync

Thunderbird: per-device account setup. Epistles: zero-knowledge Cloud Vault carries connected-account credentials between devices, encrypted under your Epistles password.

ProtonMail support

Thunderbird: through Proton Bridge, separately installed. Epistles: native, no Bridge, OpenPGP on device.

iOS support

Thunderbird: in development, TestFlight build available, public release targeted later 2026. Epistles: shipping today.

Push notifications

Thunderbird: IMAP IDLE for providers that allow it. Epistles: Gmail Pub/Sub, MS Graph subscriptions, JMAP push, IMAP IDLE for the rest.

Calendar and contacts

Thunderbird: built-in calendar with deep capability, CalDAV, address book. Epistles: CalDAV, CardDAV, JMAP, Google, Microsoft calendar and address book reading; less mature than Thunderbird today.

Pricing

Thunderbird: free, donation-supported. Epistles: free for 3 accounts; $35/yr Pro for unlimited; 15-day trial, no card.

Open source state

Thunderbird: open source since 2003. Epistles: source migration to public repo in progress; not yet open as of this page.

Telemetry

Thunderbird: privacy-respecting, no ad tracking, no AI training on conversations. Epistles: zero analytics, zero crash reporting SDKs, zero metrics; only functional API calls.

Who should pick which

If a fully open-source client today is a hard requirement, or if your mailbox is primarily desktop and your provider speaks IMAP cleanly, Thunderbird is the right choice. Twenty years of work, a built-in calendar few clients can match, an add-on catalogue, and a foundation governance model are not things a five-year-old project gets to wave away.

If iOS matters this year, or your inbox spans providers whose native protocols carry concepts IMAP cannot represent (Gmail labels, MS Graph categories, JMAP threads, Proton encryption without a separate Bridge), Epistles is the better fit today. The Cloud Vault is the other reason: signing in once across a Mac, a Linux laptop, and a phone is a quietly large daily difference.

When the source goes public, this page gets fairer

We owe an honest acknowledgement that the open-source asymmetry is real and one-sided right now. Thunderbird's source has been readable for two decades. Epistles' source is in the middle of a public migration to github.com/epistlesapp/epistles. When that lands, this page becomes a comparison between two clients you can read the code for, which is the comparison we want. Until then, take the architectural claims as claims, not proofs, and weight that accordingly.

A line at the end

Thunderbird is a project we admire and consider a sibling, not a competitor. The internet has space for more than one open, multi-account, cross-platform mail client, and it should. If your inbox fits the shape this page describes, our front door is here. If it fits Thunderbird's, theirs is at thunderbird.net, and we mean that.