Livaraprivate messengerMost comparison pages are a competitor grid with one column mysteriously full of ticks. This one asks ten questions and answers them for Livara — including the answers that count against us.
The right-hand column describes what dedicated private messengers typically answer, from their own published documentation. It is a summary of a category, not an accusation about a named product.
Channel content is readable, LGS1 membership is server-controlled, metadata is not hidden, group calls are an Android peer mesh rather than MLS/SFrame, and there is no external audit report to point at. A comparison that hid those would not be worth the scroll.
Read the full threat model ↗| The question | Livara Chat | Typical private messenger |
|---|---|---|
| Are one-to-one messages end-to-end encrypted? | Yes — current secure chats use the hybrid-PQ LVR1 ratchet, and a send is refused rather than quietly downgraded | Usually yes on dedicated messengers, usually no on social platforms |
| Are group messages end-to-end encrypted? | Private text, media, files, captions and edit content use LGS1 sender keys. Membership state is still server-controlled, so MLS-grade malicious-server integrity remains future work | Varies; coverage for text, history, media and membership changes is often not separated |
| Are channel messages end-to-end encrypted? | No. Channel content remains server-readable and is labelled separately from encrypted group text | Usually not for public or broadcast channels |
| Is the encryption resistant to a future quantum computer? | Custom LVR1 combines ML-KEM-768 with P-256 and periodically advances a PQ epoch, but it is unaudited and not equivalent to Signal's Triple Ratchet | Not comparable as a category: published post-quantum deployments now differ materially by product and protocol |
| Does the server learn your password? | No. SRP proves knowledge without the wire ever carrying the password | Commonly a password hash is sent over TLS and verified server-side |
| Is who-you-talked-to hidden from the operator? | No. Routing, timestamps and membership are stored so devices can sync | No, with rare exceptions that trade away instant multi-device sync |
| Can you verify the app you installed is the app that was published? | Partly — the SHA-256 is published and the Proof Lab hashes your file offline, but independent source-to-artifact reproduction is still missing | Store installs are signed, but a checksum you can check yourself is uncommon |
| Does a correct password alone restore your keys on a new device? | No, deliberately. The encrypted key backup needs your recovery phrase | Often yes, via a cloud backup — convenient, and a much larger blast radius |
| Is there a phone number requirement? | No. An account is a username and a password | Frequently required, which ties the account to a real-world identity |
| Is it free? | Yes, with no advertising and no message-content processing for advertising | Usually free; the funding model is what varies |
It is a question about who you are hiding from. These three answers point at three different apps, and only one of them is this one.
Livara encrypts compatible direct and private-group text/media, but its first-contact directory and LGS1 member list are still server trust points. It is the wrong choice when you require an audited protocol or malicious-server-safe group membership today.
Livara is the wrong tool. Routing, timestamps and membership are stored by the ordinary delivery path. Look at systems with integrated sealed sender, private group credentials and a published metadata threat model; multi-device support does not by itself make metadata resistance impossible.
The APK digest is published and the Proof Lab hashes your file locally, but that proves only equality with the published artifact. It does not prove source correspondence, protocol correctness or an independent audit. Those remain release gates.
The Proof Lab runs on your device and makes no network requests. It verifies selected artifact and primitive checks, not the whole security architecture.