Skip to content

Molly: Open-Source Signal Fork With App Lock

Molly: Open-Source Signal Fork With App Lock

Molly review for Android: the FOSS build without Google Play Services, with a close look at notification latency, battery tradeoffs, and SQLCipher passphrase.

App summary

App
Molly
Developer
MollyIM
License
GPL-3.0-only
Version
unknown
Size
unknown
Price
Free

Molly-FOSS Without Google Play Services: Notification Latency and Battery Tradeoffs

Molly ships in more than one shape, and the difference matters more than most reviews admit. The standard build may rely on Google/FCM for push notifications. The Molly-FOSS variant is intended to avoid Google Play Services dependencies altogether. That single sentence is the entire factual basis for the notification discussion, and it is worth reading carefully, because it says nothing about how fast messages arrive or what the battery cost is.

What the fact data does say is that feature and notification behavior can differ between the standard and FOSS variants. Users should not assume identical delivery behavior across the two. If you install Molly-FOSS expecting the same push pipeline as the Play Services build, you are making an assumption the project data does not support. The honest position is that delivery mechanics are variant-specific and need to be checked per release rather than inferred from the name on the icon.

One widespread claim deserves correction here: the idea that Molly-FOSS cannot receive push notifications at all. The fact data does not say that. It says the FOSS variant is intended to avoid Google Play Services dependencies, which is a different and narrower statement. Avoiding a dependency is not the same as having no notification path. How notifications actually reach a device without Play Services is not specified in the material available, so treat any confident answer you read elsewhere as unverified.

Battery drain is the other half of this question, and it is genuinely unknown. Exact battery impact, notification latency, permission sets, and telemetry are not specified in the fact data and must be checked for each release. That is not a dodge; it is the accurate state of the information. A fork maintained by a small team can change its notification approach between builds, and the project does not publish a stable, quotable figure in the material summarized here.

Practically, this means the variant choice should follow your device setup rather than a benchmark you read somewhere. If your phone runs Google Play Services and you want the most conventional delivery path, the standard build is the one the data describes as relying on FCM. If you have deliberately removed or avoided Play Services, Molly-FOSS is the build intended for you, and you should expect to verify notification behavior on your own hardware before relying on it for anything time-sensitive.

Migrating from Signal to Molly on Android: What the Fact Data Does and Doesn’t Confirm

The migration question is inevitable. Molly is an Android fork of Signal, built on the Signal Protocol, and it inherits Signal’s phone-number-based registration. When two apps share a protocol, a registration model, and a codebase lineage, people reasonably assume they can move an account and its history from one to the other. The fact data explains why the question arises. It does not answer it.

What is not documented is substantial: backup compatibility between Signal and Molly, whether both apps can act as primary devices on the same phone number at the same time, what the deregistration risks are, and the actual step-by-step migration procedure. The exact latest version number and the minimum Android version are also unknown. That last point matters more than it sounds, because migration instructions are version-sensitive. A procedure written for one build can be wrong or destructive on another.

On the same-number question specifically, the fact data does not confirm that Molly and Signal can both be primary devices on one number simultaneously. Treat simultaneous primary use as unverified. The registration model is inherited from Signal, which means the account is tied to a phone number rather than to an anonymous identifier, but the data stops short of describing how the two clients would coexist on that number. Do not restructure your account based on a forum post.

The sensible approach is to back up before touching anything, then verify the current procedure against the official project at molly.im and its release notes rather than against a tutorial of unknown vintage. Because Molly inherits phone-number registration, migration is an account-level operation, not an opportunity to start fresh with a new identity. If what you actually want is a messenger that does not require a phone number, migration is the wrong frame entirely, and Session is the alternative the fact data points to for that need.

Molly’s SQLCipher Passphrase: What Happens If You Forget It

Molly’s headline addition over standard Signal is local protection: an encrypted local message database using SQLCipher, plus app lock and passphrase protection. The database encryption is the substantive part. It means the message store on the device is encrypted at rest rather than sitting in a readable form that any process with file access could lift.

The question everyone asks next is what happens when the passphrase is gone. The fact data does not state whether a forgotten passphrase can be recovered, and it does not state whether encrypted database backups are recoverable. That absence is the answer this article can honestly give. There is no documented recovery path in the material, so no recovery path should be promised.

Treat the passphrase accordingly: as critical, not as a convenience setting you can reset through an email link. If the database is encrypted with a key derived from your passphrase and the project documentation does not describe a recovery mechanism, the safe assumption is that losing the passphrase means losing access to the encrypted local store. Check release-specific documentation before you rely on any assumption in either direction, because the behavior may differ between builds.

It is also worth separating the two features the fact data lists. App lock and passphrase protection are named alongside the SQLCipher database encryption, but the material does not specify precisely how the app lock relates to the database key. One may gate access to the interface while the other controls decryption, or they may be more tightly coupled. Because that relationship is unspecified, do not assume that disabling or bypassing the app lock leaves the encrypted database untouched, and do not assume the reverse either.

Molly and Signal Desktop Linking: Compatibility Limits the Facts Leave Open

Molly is an Android application. The fact data lists Android as the platform, with no iOS build and no standalone desktop client. That framing alone should temper expectations about linked devices, because the linked-device model in the Signal ecosystem depends on a desktop or secondary client talking to a primary phone. Molly’s side of that arrangement is not confirmed in the material.

Specifically, the fact data does not confirm Signal Desktop linking or linked-device compatibility for Molly. It is not a claim that linking fails. It is an absence of confirmation, which is a different thing and should be treated as such. Anyone who needs a working desktop companion should verify the current state against the official project and release notes before committing to the fork.

The variant split adds another layer. Standard and FOSS builds can differ in feature and notification behavior, so linked-device support should not be assumed to match official Signal even if one variant happens to work. A feature that functions in the Play Services build is not automatically present in Molly-FOSS, and the fact data gives no compatibility matrix to consult.

The practical upshot is narrow but important. If your daily workflow depends on reading and replying from a desktop, Molly is not the safe default. The official Signal client is the reference implementation with the larger development team and user base, and it is the one whose linked-device behavior is the baseline. Molly may suit you, but it should be adopted with the desktop question left open rather than assumed settled.

GPL-3.0-only, MollyIM, and the Politics of Maintaining a Fork

Molly is released under GPL-3.0-only and maintained under the mollyim GitHub organization, with molly.im as the official project site. It is free. Those facts do more work than they first appear to, because they define what kind of project this is: inspectable, forkable, and legally obligated to remain so in its distributed form.

The license is the reason the local database encryption and app lock can be examined rather than taken on faith. Anyone with the skills can read how the SQLCipher integration is wired and what the app lock actually gates. That is a meaningful difference from a closed-source messenger that asks you to trust a privacy claim. It is also the reason the FOSS variant can exist at all as a build that avoids Google Play Services dependencies.

What the license does not buy you is institutional backing. Molly is a smaller project with a smaller community than the official Signal app, and nothing in the fact data suggests it is an official Signal release endorsed by the Signal Foundation. It is a fork. Forks are maintained by whoever shows up, and their continuity depends on that. This is not a criticism of the maintainers; it is a description of the support model you are choosing.

That distinction matters most for people who need a vendor relationship. If your requirement is official Signal Foundation support, warranties, or a formal escalation path, Molly is not that product and will not become it. If your requirement is auditable code and a build without Play Services, the GPL-3.0-only license and the separate FOSS variant are exactly the point.

What SQLCipher and App Lock Do — and Don’t Do — for Your Threat Model

The features Molly adds are local. An encrypted local message database protects the message store on the device. App lock and passphrase protection gate access to the application. Both are aimed at the person who picks up your unlocked phone or the attacker who extracts files from storage, not at the network.

They do not change what the network sees, and they do not change how you register. Molly inherits Signal’s phone-number-based registration, so the account is still tied to a number. The local encryption is not an anonymity feature and should not be marketed as one. Anyone telling you Molly hides your phone number is describing something the fact data does not support.

Within that boundary, the additions are genuinely useful. Standard Signal does not, per the fact data, include this local database encryption and app lock layer; Molly adds them on top. For a device that travels, gets handed to a child, or sits on a desk in a shared office, an encrypted store plus a lock is a real improvement over a plaintext local database.

The limit is that the fact data does not specify the cryptographic details, key derivation, or how the app lock and database passphrase interact. Without those specifics, the right posture is to credit the feature for what it plainly is — encryption at rest plus an access gate — and not to extrapolate further. It is a hardening layer for a phone-based messenger, not a redefinition of the trust model.

Sideloading, Updates, and the Unknown Version Number

Three facts are missing that would normally anchor a review: the exact latest version number, the exact APK size, and the minimum Android version. There is also no install or download count in the material. Those gaps are not cosmetic. They mean this article cannot tell you whether your device is supported, how large the download is, or whether the build you are running is current.

The FOSS variant sharpens the problem. A build intended to avoid Google Play Services dependencies implies a distribution path that does not depend on the Play Store, which in turn implies manual installation and manual updates. The fact data does not describe how distribution works beyond naming molly.im as the official site, so that is where you should look, and nowhere else. Sideloading an APK from a mirror is a bad habit regardless of how good the app is.

Update discipline then becomes your responsibility rather than the store’s. Because the version number is unknown here, you cannot use this article to judge currency. Check the official project and its release notes on whatever cadence you think is appropriate, and treat a stale build as a real risk rather than an inconvenience.

This is the clearest audience filter in the whole assessment. People unwilling to manage APK sideloading and updates should not run Molly, and especially not the FOSS variant. The official Signal client handles that plumbing for you. Molly asks you to handle it yourself, and it does not pretend otherwise.

One Network, Two Clients: Why Molly Is Not a Separate Messenger

Molly is a fork of Signal Android that uses the Signal Protocol. It is not a separate messenger network. There is no Molly-only social graph, no parallel address space, and no walled garden where you can only talk to other Molly users. The protocol is the shared substrate, and the fork is a different client on top of it.

That distinction answers a cluster of related questions at once. Messages from Signal users are not a foreign-network problem in principle, because the underlying protocol is the same. Whether calling works identically is not enumerated in the fact data, so verify that specific feature rather than assuming. But the framing of “will Signal users be able to reach me” is the wrong worry; the right worry is variant-specific behavior and version currency.

Two misconceptions are worth killing outright. First, Molly does not make you anonymous or hide your phone number — registration is inherited from Signal and remains phone-number-based. Second, Molly does not remove the need for a phone number. If anonymous sign-up without a number is your requirement, this is not the tool, and Session is the alternative the fact data names for decentralized private messaging without a phone number.

The flip side is that Molly is not a place to build a separate community. People who want Telegram-style cloud chats and large public channels are looking at a different category of product entirely. Molly is a hardened client for private one-to-one and group messaging on an existing protocol, and it should be judged on that narrower scope.

Verdict

Molly is a hardened, open-source Signal Android fork for technically comfortable users. It earns that description through specific, verifiable additions: an encrypted local message database using SQLCipher, app lock and passphrase protection, a GPL-3.0-only license, and a Molly-FOSS variant intended to avoid Google Play Services dependencies. Those are real improvements over the stock client for the right person, and the code being inspectable is not a small thing.

It is not an anonymous messenger. Registration is still phone-number-based, and no local encryption feature changes that. It is not a separate network, so the “Molly community” framing is wrong. It is not an official Signal release endorsed by the Signal Foundation, and it is not a product with a vendor support relationship behind it. The project is smaller than official Signal, with a smaller community, and that has consequences for continuity and for how quickly variant-specific quirks get documented.

Buy-in requires accepting three tradeoffs. You take on phone-number registration because it is inherited. You take on variant-specific notification behavior, with no published latency or battery figures to guide you. And you take on sideloading and update management, particularly with the FOSS build. If any of those is a dealbreaker, the official Signal client is the correct choice: same protocol, larger development team, larger user base, and none of the update plumbing on your plate.

For decentralized or number-free requirements, the fact data points elsewhere. Element is the Matrix client for decentralized, end-to-end encrypted communication. Session is the decentralized private messenger that does not require a phone number. Neither is a drop-in Molly substitute; they are different answers to a different question.

Explicitly, Molly is not for people who need anonymous sign-up, iPhone users expecting a Molly iOS app, non-technical users who want zero-configuration messaging, anyone expecting a Molly-only social network, anyone who needs official Signal Foundation support and warranties, anyone who wants Telegram-style cloud chats and large public channels, or anyone unwilling to manage APK sideloading and updates. For the user who runs Android, understands what a fork is, wants the local database encrypted, and will check molly.im and the release notes before trusting a build, Molly is a defensible choice — provided you treat the passphrase as unrecoverable and the desktop-linking question as open until you verify it yourself.

Pros

  • Open-source and based on Signal's messaging protocol
  • Adds local database encryption and app lock features over standard Signal
  • FOSS variant available for users who want to avoid Google services

Cons

  • Inherits Signal's phone-number-based registration
  • Smaller project and community than the official Signal app
  • Feature and notification behavior can differ between standard and FOSS variants
Download from the official source

The link goes straight to MollyIM. We do not host app files.

Open official page →

Frequently asked questions

Is Molly free?

The recorded pricing model is Free.

What license does it use?

It is listed under GPL-3.0-only.

Where can I download the official version?

The official link points to https://molly.im/.

This article was drafted with AI assistance using official developer documentation, release notes, privacy policies, and public data, then reviewed by an editor. We do not claim lab testing. The full method is on our Review Method page.