CraysNostria
Nostria is not a tiny Nostr toy with a timeline and a post box. It is trying to be the broad, everyday social client: the place where feeds, messages, articles, music, streams, photos, zaps, signers and device-specific installs all feel like parts of one life.
The client that wants to give you your friends back
Nostria's best sentence is not technical. It is emotional: social without the noise, where you can see your friends again. That line tells you more about the product than a list of NIPs ever could. Nostria is built by people who think the social web went wrong when engagement became a machine for pulling you away from the people you actually care about.
The homepage and README keep returning to the same idea. Humans are social. Social media can bring people closer. But the old platform model treats users as the product, rewards manipulation and builds growth by turning attention into a resource to be extracted. Nostria wants the opposite: a social network for human connection, with user-owned data, user-controlled profiles and a Nostr-based identity that cannot be owned by the app.
That is why Nostria feels different from smaller Nostr clients. It is not only asking, "Can we show notes?" It is asking whether a Nostr client can become a real social home: one app where you post, follow, read, watch, listen, zap, write articles, join streams, manage media and keep enough control that leaving the app does not mean losing yourself.
This makes Nostria both exciting and risky. A narrow client can be judged by one thing: does it post and read well? Nostria has chosen the harder route. It wants breadth. That means you have to judge it like a product ecosystem, not like a timeline widget.
Who is building it
The public project sits under the GitHub organization nostria-app. The main repository is public, MIT-licensed and described as "Nostria: Your Social Network. Built for human connections. See your friends again. Nostria is social without the noise." When checked for this article, the repository had 53 stars, 12 forks, 98 open issues, TypeScript as its main language, a first public creation date of April 6, 2025, and a very recent push on June 6, 2026.
The official press trail names Sondre Bjellas as creator, with Brainbox as the company behind Nostria. One early release announcement says Nostria came out of seven months of development and lists features such as cross-platform access, local models, Blossom-based decentralized media storage, unlimited events, live streaming, articles, Nostr Wallet Connect, zaps and badge workflows. Another funding announcement says Nostria raised a pre-seed round through Angor's Bitcoin-native crowdfunding protocol.
There is a useful transparency detail here: Nostria does not hide behind a dead landing page. It has a public repo, public docs, public press posts, releases, a status page, GitHub issues, a separate source-code architecture document and a real download story. That does not make every product decision correct, but it gives the reader something to inspect.
The funding story is also very Nostr-adjacent. Angor describes itself as Bitcoin-based, on-chain and milestone-oriented. Nostria's own seed-fund material frames the product as a client that can bring Nostr to broader adoption by making it easier, faster and more complete. You can feel the ambition: not merely "a Nostr client exists", but "a normal person should be able to live here."
One app, many rooms
Open the public site and you see the product trying to speak to several kinds of users at once. The hero line is human connection. The feature pages talk about creators, musicians, live streamers, writers, photographers and short-video makers. That could become messy, but the logic is consistent: Nostr is a protocol for signed social objects, and Nostria wants to make many of those objects feel native.
For ordinary social reading, Nostria offers feeds, following, trending, search and customizable views. The NostrApps entry highlights multiple feeds with multiple columns and an outbox model. The architecture document describes feeds with multiple sources such as following, public, trending and search, plus account-level preferences and per-account local state. That gives Nostria a dashboard flavor: less like a single endless scroll, more like a workbench for people who actually want to watch different parts of the network.
For writers, Nostria supports long-form articles, the NIP-23 article flow and a public site section that explicitly talks about publishing essays, updates and newsletters from one identity. It even names other Nostr readers such as YakiHonne, Habla, Highlighter, Primal, Damus and Amethyst as places where published writing can travel. That is the right instinct: writing on Nostr is not valuable because one app has a pretty editor, but because the article can become part of a wider reading network.
For media creators, Nostria goes further. The site talks about music, streams, short clips, photo libraries and zaps. The architecture document lists a media player system, live streaming viewer, music player with offline support, video events, voice messages and Blossom file storage. That is a serious spread. If implemented well, Nostria becomes a kind of Nostr studio. If implemented badly, it risks becoming a heavy app with too many doors. The current public code shows the ambition is real; the user experience has to make that ambition feel calm.
Under the hood
Nostria is built as an Angular application, with Angular Material, TypeScript, signals, nostr-tools and a large dependency set around media, wallets, local storage and browser-side processing. The package file shows Angular 22 packages, nostr-tools, @getalby/sdk, @cashu/cashu-ts, @huggingface/transformers, hls.js, mediabunny, music-metadata, marked, dompurify, idb and Tauri packages. That is not a toy stack.
The desktop story is Tauri. The Tauri config identifies the product as Nostria, the app identifier as app.nostria, the bundle publisher as Brainbox and the category as SocialNetworking. It targets AppImage, deb, dmg, NSIS and MSI packages, has deep-link schemes for nostr and nostr+walletconnect, and uses a GitHub releases endpoint for updates. The same repository also contains Android Tauri and Android TWA work, plus web/PWA handling.
This cross-platform strategy is not cosmetic. Nostr clients live or die by key handling, notification reliability, media permissions and install habits. A browser app is easy to open. A desktop app can feel more stable and private. Mobile apps need signer integration, notifications and camera/microphone flows. Nostria is trying to cover all of that from one codebase while still shipping native wrappers where the platform demands it.
The backend-service layer is also explicit. The architecture document lists an image proxy, metadata service, discovery relay and CORS proxy using Nostria domain patterns. That is a subtle but important point. Nostr is decentralized, but a polished app often adds services around it for media optimization, previews, relay discovery and startup speed. Nostria is honest enough to document those services. The reader should treat them as convenience infrastructure, not as the Nostr protocol itself.
Nostria has opinions
The README calls Nostria an opinionated Nostr client, and that is not a throwaway line. It says Nostria ignores NIP-65 read/write relay flags and treats all relays as both read and write. It says it does not support NIP-96 HTTP file storage because it prefers Blossom. It says badges should be self-contained on the user's relays, so both badge definition and badge claim are published there.
You may agree or disagree with those choices. The important thing is that they are stated. Nostr is not a single product spec with one final user experience. Every serious client makes decisions about relay lists, outbox discovery, file storage, badges, labels, moderation, search, media, sync and private messages. A client that pretends to be neutral often hides its opinions in confusing behavior. Nostria writes some of them down.
The supported-NIP list is long. The README includes NIP-01, NIP-02, NIP-04, NIP-05, NIP-06, NIP-07, NIP-17, NIP-23, NIP-44, NIP-46, NIP-47, NIP-50, NIP-51, NIP-53, NIP-57, NIP-58, NIP-65, NIP-66, NIP-68, NIP-71, NIP-75, NIP-78, NIP-85, NIP-88, NIP-98, NIP-A0, NIP-B7 and Blossom BUDs, among others. The architecture document is slightly shorter in its table but points to the same broad implementation territory.
That breadth is why Nostria deserves a real test instead of a quick install-and-forget impression. A client can list many NIPs and still make one of them awkward. The interesting question is how those standards feel together. Does private messaging feel safe? Does Blossom media upload feel natural? Does outbox behavior make posts easier to find? Do long-form articles travel well? Does Nostr Wallet Connect feel like a native action rather than a separate wallet chore?
Local models, media and zaps
Nostria's local model features need a careful sentence because this is where many apps slide into noise. The interesting claim is not "AI". The interesting claim is locality. The architecture document says Nostria uses Transformers.js in a web worker, with models cached in IndexedDB, and that user content does not leave the device for those tasks. It lists summarization, translation, transcription, text-to-speech, sentiment analysis and text generation as supported task categories.
Used well, that can become a humane tool. A long thread can be summarized before you sink half an hour into it. A post in another language can be translated without shipping the text to a third-party API. A voice command can be transcribed locally. Used badly, it can become decoration. The key is whether Nostria keeps these tools quiet, optional and device-local. The docs say the settings layer has separate toggles for summarization, translation, transcription, speech and sentiment, which is exactly the kind of control a privacy-minded client needs.
Media is the second big layer. Nostria uses Blossom rather than NIP-96, supports local compression and server compression options, and uses Mediabunny for browser-side media processing. The architecture document describes image compression through canvas and video compression/transcoding in-browser before upload. That matters because media is where many Nostr clients become clumsy. A social client without good media handling starts to feel like a protocol demo.
Then there are zaps. Nostria supports NIP-57 Lightning zaps and integrates @getalby/sdk for wallet work. The site talks about creators receiving value directly from audiences. This is one of the places where Nostria's broad approach makes sense: if a writer publishes an article, a musician shares a track, a streamer goes live and a photographer posts a gallery, all of them should be able to receive value without leaving the social surface.
The companion signer is another piece of the trust story. The public site promotes Nostria Signer as a Chrome extension for secure Nostr signing, usable with Nostria or other clients. That is good product thinking: a big social client should not normalize casual private-key handling. The question to test is whether Nostria makes signer-based use easy enough that ordinary users do the safer thing by default.
What to test before you move in
If you are only curious, open the web app and read. If you are considering Nostria as a home client, test it like a home. Start with identity. Can you use an existing key through a signer? Are permissions clear? Can you switch accounts without confusion? Does the app treat a Nostr user as a Nostr user, not as a Nostria-only account?
Then test relay behavior. Nostria says it is built around scalability, outbox thinking and relay discovery. Create a normal post, a reply and a long-form article. Check whether other clients can find them. Look at where they were published. If an event matters to you, do not just trust that it appears in Nostria's own feed.
Test media with small files first. Upload an image, a short clip and maybe an audio post. See where the files live, whether Blossom server choices are visible, whether compression changes quality, and whether another client can open the result. Nostria's media stack is one of its strongest claims, so it deserves real hands-on pressure.
Finally, test weight. A broad client can become the app you keep open every day, or it can become a menu maze. Nostria's promise is that a richer Nostr client can still be human. The product wins when the feed feels calmer, the writing tools feel discoverable, the media tools feel practical and the payment/signing parts do not yank you out of the moment.
The larger point is this: Nostria is not valuable merely because it implements many NIPs. It is valuable if those protocol pieces help you see people again, publish without surrendering your identity, carry your media across apps and receive value directly. That is a high bar. Nostria is one of the few clients openly trying to clear it.
Sources worth opening
This article keeps claims close to the public project trail. Start with these pages when you want to verify the product yourself.
- Nostria official site
- Nostria features page
- Nostria about / manifesto page
- Nostria MVP release announcement
- Nostria pre-seed funding announcement
- Nostria seed fund launch announcement
- Nostria documentation
- Nostria quick start documentation
- Nostria FAQ
- Nostria GitHub repository
- Nostria README
- Nostria package.json
- Nostria architecture document
- Nostria Tauri config
- Nostria releases
- GitHub repository metadata API
- NostrApps listing for Nostria
- Nostria Signer on Chrome Web Store
- NIP-25 reactions
- NIP-27 text note references
- Nostr protocol NIPs repository
- NIP-01 basic protocol flow
- NIP-05 DNS identifiers
- NIP-07 browser signer capability
- NIP-10 text note threading
