Threema launched in 2012 and is developed by a company based in Switzerland. Its pitch has stayed consistent: end-to-end encrypted messaging that collects as little metadata as possible, sold as a paid app rather than a free service supported by advertising. That business model is part of the privacy argument — when you're the paying customer, you're not the product.
The Headline Feature: No Phone Number
Threema's most distinctive choice is that you don't need a phone number or email address to sign up. Instead, the app generates a random Threema ID — an eight-character string tied to a key pair created on your device. You can share that ID, a QR code, or a scannable link to let others add you. Providing a phone number or email is optional, used only to help contacts find you, and the app stores those identifiers as hashes rather than in the clear.
This directly addresses the phone-number problem that affects identifier-based messengers: your account isn't anchored to a number that can be SIM-swapped, subpoenaed, or used to link your messaging identity to your real-world one. For activists, journalists, or anyone who simply doesn't want their chat identity tied to a phone, that's a real advantage.
Because your identity is a key pair on your device, Threema's security — like any key-based system — depends on verifying contacts. Threema uses a three-dot trust indicator (red, orange, green) that goes green only when you've scanned a contact's QR code in person. Until then, you're trusting the server's key directory, the same trust-on-first-use assumption most messengers make.
The Cryptography and the Open-Source Question
Threema's encryption is built on the well-regarded NaCl / libsodium primitives rather than hand-rolled cryptography — a good sign, since standing on vetted primitives is exactly what you want. For years a common criticism was that the apps were closed-source, which makes independent verification harder. Threema addressed this in late 2020 by open-sourcing its client apps, so the code that runs on your device can now be inspected. The server remains proprietary, which is the norm for commercial messengers.
The company has commissioned external security audits and publishes them, and it has a published forward-secrecy-supporting protocol that it rolled out to strengthen the guarantees of its connection layer over the older design.
The 2023 Academic Findings
An honest review has to include this. In early 2023, researchers in the Applied Cryptography group at ETH Zurich published an analysis of Threema's protocol that identified seven distinct weaknesses across various attack models — including issues around message ordering, replay, and certain cross-protocol interactions. The work was rigorous and is exactly the kind of scrutiny a security product should welcome.
Threema's response was measured: it argued that several of the issues required strong preconditions (such as a compromised server or substantial user error), that some were already being addressed by the newer protocol it had been rolling out, and that it had no evidence of real-world exploitation. Both things can be true at once — the findings were legitimate and worth fixing, and they did not amount to "Threema is broken, your messages are exposed."
The right lesson from an academic audit isn't "this app failed." It's that protocols benefit from adversarial review, and a vendor that responds by hardening its design is behaving the way you want a security vendor to behave. — On reading security-research headlines
What it does mean for you: Threema's security is real but not infinite, the protocol has evolved in response to scrutiny, and — as with any messenger — keeping the app updated matters, because fixes only protect you once you're running them.
The Trade-offs Worth Naming
| Strength | Counterweight |
|---|---|
| No phone number required | Smaller network — fewer of your contacts are likely already on it |
| Paid app, no ad model | The upfront cost is a barrier to the casual contacts you want to reach |
| Swiss jurisdiction, minimal metadata | Swiss law still permits lawful compulsion for the limited data the service holds |
| Open-source clients (since 2020) | Server remains closed; you trust the operator for server-side behavior |
| Audited and actively maintained | 2023 research found protocol weaknesses — fixed/mitigated, but a reminder nothing is bulletproof |
The network-size point is the one most people underestimate. The best encrypted messenger is the one your contacts will actually use, and a paid app with no phone-number onboarding asks more of the people you want to reach. Threema partly answers this with Threema Work, a business edition that organizations deploy to their whole staff — which is arguably where the app is strongest, since the network problem disappears inside a company.
Who Threema Is Actually For
Threema is a strong fit if your priority is not linking your messaging identity to a phone number, you're comfortable paying for software, and you can get your key contacts onto it. It's especially compelling for organizations that want a Swiss, low-metadata, centrally deployable messenger for staff communication.
It's a weaker fit if you need to reach a broad set of casual contacts who won't pay or install another app, or if your communication needs extend past chat into encrypted email and documents — Threema is a messenger, not a full communication suite.
How Haven Compares
Haven shares Threema's instincts — minimal metadata, vetted primitives, no business model that depends on monetizing you — and extends them in two directions. First, messaging uses the MLS protocol (RFC 9420), an IETF-standardized group-messaging protocol with formal forward-secrecy and membership-change guarantees. Second, Haven combines encrypted chat and encrypted email under a single identity, so you're not maintaining one app for messaging and another stack for mail.
None of that makes Threema a wrong choice. It's a well-built app with a clear philosophy, and for the no-phone-number use case it's one of the genuinely good options. The point of a review like this isn't to crown a winner — it's to make sure you're choosing on the basis of how each tool's real threat model lines up with yours. If you want a deeper side-by-side, our messenger comparison lays out the broader landscape.