Composer 2 scores highest in React Native evals with 96%

react-native-maps is the de-facto standard but built on Objective-C&#x2F;Java with the legacy bridge.I rewrote it from scratch in Swift and Kotlin using Nitro Modules.Direct wrapper of the Google Maps SDKs on both platforms — no Apple Maps fallback.Full SDK coverage: markers (including SVG), polylines, polygons, circles, heatmaps, KML layers, tile overlays, traffic, transit, 3D buildings, indoor maps, dark mode, snapshots, and Street View (beta).GitHub: <a href="https:&#x2F;&#x2F;github.com&#x2F;pinpong&#x2F;react-native-google-maps-plus" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;pinpong&#x2F;react-native-google-maps-plus</a>Docs: <a href="https:&#x2F;&#x2F;pinpong.github.io&#x2F;react-native-google-maps-plus" rel="nofollow">https:&#x2F;&#x2F;pinpong.github.io&#x2F;react-native-google-maps-plus</a>

Show HN: React-native-Google-maps-plus – Google Maps for React Native

On-Device AI Coming to React Native with Gemma and React Native Executorch

We built browser-metro, a Metro-like bundler that runs entirely in a Web Worker. It supports full HMR with React Refresh, Expo Router with file-based routing, and on-demand npm package resolution via an ESM server. API routes run in-browser through fetch interception — no server or service worker needed.
Unlike Expo Snack (server-side bundling) or CodeSandbox, everything here happens client-side. Currently web-preview only; native device preview is on the roadmap.Open source (MIT): <a href="https:&#x2F;&#x2F;github.com&#x2F;RapidNative&#x2F;reactnative-run" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;RapidNative&#x2F;reactnative-run</a>

Show HN: ReactNative.run – Browser Metro bundler that runs React Native

Thinking of harness on mobile? I ported OpenAI Agents SDK to React Native

The Complete Guide to React Native Build Optimization

A floating performance overlay for React Native. See your app&#x27;s FPS, memory, re-renders, and network activity in real time, right on screen.

Show HN: React-native-sightline – floating dev overlay for FPS/memory/re-render

Hi folks, today we&#x27;re launching QVAC SDK [0], a universal JavaScript&#x2F;TypeScript SDK for building local AI applications across desktop and mobile.The project is fully open source under the Apache 2.0 license. Our goal is to make it easier for developers to build useful local-first AI apps without having to stitch together a lot of different engines, runtimes, and platform-specific integrations. Under the hood, the SDK is built on top of QVAC Fabric [1], our cross-platform inference and fine-tuning engine.QVAC SDK uses Bare [2], a lightweight cross-platform JavaScript runtime that is part of the Pear ecosystem [3]. It can be used as a worker pretty much anywhere, with built-in tooling for Node, Bun and React Native (Hermes).A few things it supports today:<pre><code> - Local inference across desktop, mobile and servers
 - Support for LLMs, OCR, translation, transcription, 
 text-to-speech, and vision models
 - Peer-to-peer model distribution over the Holepunch stack [4],
 in a way that is similar to BitTorrent, where anyone can become a seeder
 - Plugin-based architecture, so new engines and model types can be added easily
 - Fully peer-to-peer delegated inference
</code></pre>
We also put a lot of effort into documentation [5]. The docs are structured to be readable by both humans and AI coding tools, so in practice you can often get pretty far with your favorite coding assistant very quickly.A few things we know still need work:<pre><code> - Bundle sizes are larger than we want right now because the current packaging of Bare add-ons is not as efficient as it should be yet
 - Plugin workflow can be simpler
 - Tree-shaking is already possible, but at the moment it still requires a CLI step, and we&#x27;d like to make that more automatic and better integrated into the build process
</code></pre>
This launch is only the beginning. We want to help people build local AI at a much larger scale. Any feedback is truly appreciated! Full vision is available on the official website [6].References:[0] SDK: <a href="http:&#x2F;&#x2F;qvac.tether.io&#x2F;dev&#x2F;sdk" rel="nofollow">http:&#x2F;&#x2F;qvac.tether.io&#x2F;dev&#x2F;sdk</a>[1] QVAC Fabric: <a href="https:&#x2F;&#x2F;github.com&#x2F;tetherto&#x2F;qvac-fabric-llm.cpp" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;tetherto&#x2F;qvac-fabric-llm.cpp</a>[2] Bare: <a href="https:&#x2F;&#x2F;bare.pears.com" rel="nofollow">https:&#x2F;&#x2F;bare.pears.com</a>[3] Pear Runtime: <a href="https:&#x2F;&#x2F;pears.com" rel="nofollow">https:&#x2F;&#x2F;pears.com</a>[4] Holepunch: <a href="https:&#x2F;&#x2F;holepunch.to" rel="nofollow">https:&#x2F;&#x2F;holepunch.to</a>[5] Docs: <a href="https:&#x2F;&#x2F;docs.qvac.tether.io" rel="nofollow">https:&#x2F;&#x2F;docs.qvac.tether.io</a>[6] Website: <a href="https:&#x2F;&#x2F;qvac.tether.io" rel="nofollow">https:&#x2F;&#x2F;qvac.tether.io</a>

Show HN: QVAC SDK, a universal JavaScript SDK for building local AI applications

Itsumo is a language learning app that generates short lessons from stories about topics you choose, at your level. Instead of working through a fixed curriculum or another flashcard deck, you can make a lesson about whatever happens to be interesting to you that day, with audio, tappable words, and grammar notes.You can try a language story from the landing page without signing up, and if you want to try full access, no credit card is required at signup, and the premium tier free for 14 days. It supports Japanese, Polish, Spanish, English, Portuguese, French and Italian.I built it because I kept bouncing between two kinds of Japanese study tools: textbooks I liked but never had with me, and apps I always had with me but didn&#x27;t actually want to open. What I wanted was something that felt closer to reading a lesson in a book than doing a drill, and I wanted it on my phone.There aren&#x27;t a lot of apps out there targeting intermediate language-learners. I&#x27;ve been studying Japanese, and a lot of the tools are optimized either for complete beginners or for spaced-repetition-heavy review. That works for some people, but once I was past the basics, I wanted to practice in the context of a story that felt relevant to me.Technical details, since this is HN:- It&#x27;s an Expo React Native app, and we&#x27;re going for universal access. It&#x27;s available on the web now, with mobile versions launching soon. 
- Created with a cool custom mobile studio I talked about here: <a href="https:&#x2F;&#x2F;youtu.be&#x2F;ufgYlZCfRM4?t=643" rel="nofollow">https:&#x2F;&#x2F;youtu.be&#x2F;ufgYlZCfRM4?t=643</a>
- Lesson generation goes through an LLM pipeline with a verification pass to keep the output near the requested JLPT level.I&#x27;m especially interested in feedback from people who are somewhere in the middle with Japanese, where beginner material feels too basic but native material is still too much work. If you try it, I&#x27;d like to hear where it feels off, confusing, or not actually helpful.Give it try it: <a href="https:&#x2F;&#x2F;itsumo.study" rel="nofollow">https:&#x2F;&#x2F;itsumo.study</a>Built by Margaret and Raf <a href="https:&#x2F;&#x2F;okthink.ai&#x2F;" rel="nofollow">https:&#x2F;&#x2F;okthink.ai&#x2F;</a>

Show HN: Itsumo: Make yourself fun and interesting academic language lessons

Hey HN. I&#x27;m Isaac, and I&#x27;m building Marco (<a href="https:&#x2F;&#x2F;marcoapp.io" rel="nofollow">https:&#x2F;&#x2F;marcoapp.io</a>), a cross-platform email client that works with any IMAP provider. macOS, iOS, and web today.I started Marco because I finally lost patience with Apple Mail, and the email client market has a weird gap. Legacy clients look terrible and&#x2F;or are not cross-platform. The good ones scan your data or cost $300+&#x2F;year. And there&#x27;s a graveyard of startups (Tempo, Big Mail, Caley) who built beautiful products and shut down after a year or two.I made a few opinionated bets early on:1. IMAP-first, not Gmail API-first. Nearly every email startup builds on the Gmail API. It&#x27;s convenient, but it locks you into one provider. Marco is IMAP-native, which means it works with Gmail, Outlook, iCloud, Fastmail, custom domains, and any provider that supports IMAP.2. Offline-first. You should be able to read, reply, delete, and organise email on a plane with no wifi. When you reconnect, everything syncs. This requirement nearly killed me. I went through WatermelonDB, Triplit, InstantDB, PowerSync, and Replicache before landing on my current approach: regular HTTP endpoints with TanStack DB and TanStack Query, using IndexedDB on web and SQLite on mobile as storage layers. I ditched &quot;fully fledged&quot; sync engines entirely. Turns out, for my data volumes (hundreds of thousands of entities per user), every sync engine I tried either choked on performance or added complexity I didn&#x27;t need. I wrote about this journey in detail: <a href="https:&#x2F;&#x2F;marcoapp.io&#x2F;blog&#x2F;offline-first-landscape" rel="nofollow">https:&#x2F;&#x2F;marcoapp.io&#x2F;blog&#x2F;offline-first-landscape</a>3. No AI. This is intentional. Every email client launching right now leads with AI. I think most of it is noise that none of us want or need. Marco is a tool. It should be fast, reliable, and stay out of your way. No email summarisation, no smart replies, no &quot;AI powered&quot; anything.The stack is React Native with Expo, Node.js backend on Railway, Postgres, Redis, S3, etc (all privately networked). Yes, a lightweight backend is needed to facilitate things like push notifications. One codebase across all frontend platforms, 100% shared code.Marco is bootstrapped and profitable at $8&#x2F;month with a 7-day free trial. 2,000+ users, all organic. No VC, no paid marketing.Would love feedback from HN. Happy to go deep on IMAP internals, the offline-first landscape, or any of the technical decisions.

Show HN: Marco, a privacy-first, offline-first email client (IMAP-native, no AI)

Hi HN, I&#x27;m an indie dev who built this to settle a long-running argument with my teenage son about who lifts more at the gym. 
I wanted a workout tracker to kept our data side-by-side on a leaderboard, and let us swap training plans without texting screenshots back and forth. 
Nothing I tried did all three without a subscription wall, so I wrote my own.Most of the app is free, it covers the full workout experience and progress tracking with 80+ exercises.
The one-time purchase unlocks multiple plans, custom movements, and the social&#x2F;sharing features, mainly to pay for the infra.The mobile app is React Native via Expo with Expo Router, NativeWind for styling, Zustand for state (every store is user-scoped via an AsyncStorage key prefix so account switches do a clean rehydrate, good way to have several using using the same app), Reanimated for animations, and react-i18next across six locales. 
The core workout loop is fully offline, and a Firebase layer sits on top for optional cross-device sync and the social features. 
Backend is a small workspace of Firebase Cloud Functions v2 that handle cross-user fan-out (friend requests, shared plans, shared movements) with rate limiting and anti-spam in Firestore triggers.Published a few weeks ago, I can proudly but surprisingly see the daily usage increasing.This is my first mobile app since a while (was doing some 10+y ago), keen for any feedback since it gave me the excitement of building some more.Android only for now, iOS to be published soon.

Show HN: Personal Trainer – Lift, share plans, and rank with friends

I&#x27;m a Senior Engineer in Berlin. My hometown is Guardia Sanframondi — a hilltop village of ~5,000 people in Southern Italy&#x27;s Sannio wine region. Local farmers grow heritage grapes (Falanghina, Aglianico) and olive cultivars that have been here for centuries.One evening I was looking at Copernicus satellite data and thought: could I turn NDVI imagery and weather data into something useful for the farmers I grew up with? That turned into months of evenings and weekends.What it does: Guardia del Campo combines satellite imagery (Sentinel-2 NDVI, soil moisture), weather data (Open-Meteo, ARPA Campania), and ML models trained on local varieties to produce daily risk forecasts for downy mildew, powdery mildew, olive fly, frost, and heat stress. Every morning, a Gemini AI brief tells farmers what to do — in plain Italian. No &quot;probability coefficient.&quot; Just: &quot;humidity is high, check the vines for peronospora today.&quot;The part I&#x27;m most proud of: anonymous crowdsourced reporting. A farmer spots disease, taps three buttons, every grower within 10km gets an alert. No account, no login, no personal data. Works offline, syncs when connectivity returns.
Stack: Next.js + Vercel (dashboard + cron jobs for ML pipeline), Supabase (Postgres + Edge Functions), Python&#x2F;scikit-learn (daily-retrained SGD models), Gemini API (Italian agronomic briefs), Capacitor (wraps the web app as native Android — same codebase, Play Store, native push). Capacitor over React Native because the entire product is a web app. No rewrite needed.Cold-start problem: Crowdsourced reporting only works if farmers report. My approach: personal emails in Italian to local producers. No marketing speak. Just: &quot;I built this, it&#x27;s free, try one report next time you&#x27;re in the vineyard.&quot; Being honest about the rough interface actually helped — farmers appreciated someone not trying to sell them something.
The data moat: Every photo a farmer uploads becomes training data for the community&#x27;s diagnostic model. After 2-3 seasons with 200+ farms, this creates a phytopathological dataset specific to Sannio varieties and microclimate. No big AgTech company can replicate this without years of local presence.The real product isn&#x27;t the app — it&#x27;s the dataset.
Economics: Revenue: €0, by design. Cost: ~€80&#x2F;month (Vercel, Supabase, Gemini API). Monetization kills trust when you have 0 users.What&#x27;s next: Vendemmia 2026 pilot with 20-30 farms.Web: https:&#x2F;&#x2F;guardia.sann.io&#x2F;campo
Play Store: https:&#x2F;&#x2F;play.google.com&#x2F;store&#x2F;apps&#x2F;details?id=io.guardiasannio.campo

I spent 2 months building a free AgTech app for my 5k-person hometown

I built an expo-module that makes it easy to integrate CallKit&#x2F;Core-Telecom into a React Native app. I&#x27;ve built a VoIP app before and this was one of the hardest parts. The existing react-native-callkeep is harder to use, isn&#x27;t easy to set up VoIP notifications with and hasn&#x27;t been updated in two years. This new module is tested with the latest versions of iOS&#x2F;Android, Expo and LiveKit. It has a simple and unified API across iOS&#x2F;Android.

Show HN: expo-callkit-telecom – easily integrate CallKit/Core-Telecom

I built MySigner because every mobile release meant juggling Xcode, Fastlane, App Store Connect, the Play Console, plus a $200&#x2F;mo ASO tool and another for reviews. Five things. I wanted one command.So the main piece is `mysigner ship appstore`. It builds, signs, uploads, waits for Apple&#x27;s processing, then submits for review. Same idea for Android: `mysigner ship play-production`. It figures out what kind of project you have on its own (native, React Native, Flutter, Capacitor, Expo). No config file unless your setup is weird.The dashboard does the rest in one place. Screenshot Studio with device frames and translated captions for 30 locales, keyword tracking, reviews from both stores in one inbox, analytics synced daily from ASC and Play.On trust because I know this matters: the build runs on your machine, not mine. Your Apple cert and provisioning profile decrypt only inside your local build, my servers never see them in the clear. Credentials at rest are AES-256. Hosted on Hetzner US, no third-party data brokers. Audit log is hash-chained on the Team plan.Stack is Rails 8 with Hotwire and Tailwind. CLI is a Ruby gem, install with `gem install mysigner`.Pricing: free tier ships to TestFlight unlimited, full CLI, 1 app. Pro is $12&#x2F;mo for 10 apps with ASO and reviews and analytics. Team is $49&#x2F;mo with SSO and audit log. 14-day Pro trial, no card.Platform-wise, iOS shipping requires macOS because Xcode is macOS-only, that&#x27;s Apple&#x27;s constraint not mine. Android works on macOS, Linux, and Windows. The CLI has no native extensions so `gem install mysigner` works wherever Ruby 3.2+ does. I&#x27;ve mostly tested on macOS though. If you try it on Linux or Windows I&#x27;d really appreciate hearing if anything broke.Fastlane is honestly still the right tool if you&#x27;ve spent months tuning your Fastfile or need self-hosted CI in your own VPC. This is for indies and small teams who&#x27;d rather type one command than maintain a release pipeline.I could really use feedback. Bugs, weird workflow stuff, anything that doesn&#x27;t make sense. Limited time and limited team so any criticism is gold.Thank you :)

Show HN: MySigner – Ship iOS and Android apps with one command

A few weeks ago I posted about Red Grid MGRS, a free 10-digit MGRS navigator for iOS (Android shipping soon!). Based on your feedback I shipped a bunch of new updates:Topographic map layer with contour lines (OpenTopoMap) alongside standard and dark tactical tilesMeshtastic LoRa mesh integration — connect to a $20 radio over BLE for off-grid position sharing. Had to write a minimal protobuf encoder from scratch since there&#x27;s no React Native SDKThree map styles you can cycle through and download for offline useFixed the iOS BLE state machine issue where CoreBluetooth starts as &quot;Unknown&quot; and silently blocks scanningGetting Meshtastic integration to work was a rabbi thole, but by far the most interesting technically. The protocol uses protobuf over BLE with a specific handshake. There was no documentation for third-party integration, so I worked from the protobuf definitions and the official iOS app&#x27;s behavior to get this working.Still zero network calls for core functionality. Still no tracking. Open source.

Show HN: Topo maps and Meshtastic mesh networking in MGRS navigator

Hello, I have made an app using React Native as I&#x27;m a solo developer and want to provide my app both on the Google Play Store and Apple Store.My app was rejected because according to Apple:&gt; We understand that you&#x27;ve provided both premium and non-premium testing accounts. Regarding 3.1.1, since the app allows users to purchase premium plan using payment mechanisms other than In-App Purchase, it isn&#x27;t in compliant with the 3.1.1 Guidelines.&gt; To resolve the issue, it would be appropriate to add the ability to pay for them in the app, using In-App Purchase.With my app currently, the user installs it and can use it for free, however it has premium features, which you unlock by subscribing to a monthly subscription. You do this by having an account on the app, and pressing a &quot;subscribe&quot; button, which then on the following page opens your browser with a Stripe payment link. You complete the checkout, and go back to the app, which registers from the backend that you have gone premium.So I&#x27;m just trying to understand, because their comment on the rejection isn&#x27;t very informative and I&#x27;m new to this: Does my app HAS to have Apple&#x27;s In-App purchasing and checkout implementation, no matter what? Or can I do it with having for example a direct link or button leading to a checkout page from Stripe, or a WebView that is still inside the app, or any other way of doing it?By the way I&#x27;m not native English.Thanks in advance!

Ask HN: Regarding app rejection on 3.1.1 Guidelines

Hello everyone, I have been working on this for a year and it is such a difficult problem. mobile-ink is an open-source mobile infinite drawing canvas. There is not a strong open-source drawing engine beyond toy demos. Thus, mobile-ink aims to solve this problem and provide a state-of-the-art mobile note-taking canvas package.mobile-ink uses native Skia&#x2F;Metal to address issues with native rendering, pencil latency, eraser behavior, page virtualization, preview caching, memory stability, page interactivity, etc. It has many advanced features such as shape-snapping, primitives to integrate figures, custom backgrounds, pdf import, stroke selection, smooth zoom&#x2F;scroll, and more.The package is iOS + React Native as of now.

Show HN: Mobile-ink, an open-source mobile infinite drawing canvas

I&#x27;m a first year CS student at Sheffield and built this over the past year because I thought I could do better than Trolley.Stack: React Native &amp; Typesense for vector search and filtering. An ETL pipeline runs in the background handling scraping and automatic product categorisation using embeddings.Covers Aldi, Tesco, Sainsbury&#x27;s, Asda, Morrisons, Waitrose, Iceland, Ocado &#x2F; M&amp;S and Co-op. Clubcard and Nectar prices included. Sort by price per unit, filter by category and stores near you.This was fun to make, and I have a lot planned, with some WIP:- Basket Generation based on entering a shopping list and getting cheapest products for each store, and cheapest collection of stores for the cheapest total shop. Feels quite magical even in dev mode right now- Local price estimates. Local stores differ from online prices by up to 20% depending on store fascia. I&#x27;m thinking to record a basket of goods for each local store and use it to estimate local prices- Crowdsourced prices. Tricky one, due to running from a cold start. Would be very useful for certain stores without online prices, eg. Lidl.That&#x27;s all for now! Hopefully it&#x27;ll help me get an internship even without traction ^_^

Show HN: I built an app for comparing grocery prices across UK supermarkets

Hi, I&#x27;m Andrew, a long-time hacker about to turn 50 in a few days. Today I&#x27;d like to tell you how taout.tv came into my life. It stands for &quot;talk it out&quot; and I&#x27;m told this is a &quot;portmanteau&quot; when you push two words together like brunch or smog. I met Greg at a Los Angeles tech networking group and he told me the idea.His pitch was &quot;I&#x27;ve literally heard Joe Rogan explain my product. Two people can&#x27;t agree on a subject, both think they are right. They should just talk it out.&quot; I agreed to a two-week trial as his CTO. He had been working on this for 2 years with an off-shore team. They picked nextjs for web frontend and backend, react native for ios and android, elasticsearch, redis, and postgres. Hosted on 8 digital ocean droplets and using 100ms.live for the live broadcast rooms.I was able to get the droplet count (and size) down to just 2, using only postgres (pgvector search), frontend web is just js react no nextjs. Backend is now all golang. And the two mobile apps are native swift and kotlin.There&#x27;s actually a lot to this platform. If you are familiar with orgs like bridgeusa.org or braverangels.org this idea of bridging people on two different sides is something a lot of people care about. You schedule a podcast with someone you already know, or you can just post your desire to talk about a subject with anyone. You and your co-host go live and the audience can watch and score each side in real time. You can upload evidence to support your side and all these scores are tracked and recorded.Would love some feedback from users with podcasting experience.<a href="https:&#x2F;&#x2F;www.taout.tv&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.taout.tv&#x2F;</a>Thanks!

Show HN: Bridging People Together via Podcasts

Hi HN, I&#x27;m William.ember is a matchmaking app I built because after 10 years of
using every major dating app, I&#x27;d close each one feeling worse
than when I opened it. To me they all look the same and feel
the same. The apps reward looks over substance and swipes over
conversations. I crave deeper connection and wanted to see if
a personality-first format can work.It&#x27;s a &quot;Love Is Blind&quot; structure: you do an interactive
onboarding that builds a persona, get compared automatically
against other users, and connect through conversation before
ever seeing a photo. Photos unlock after you&#x27;ve actually matched.A few things I tried to solve that bother me about existing apps:- Ghosting: a common issue where people just stop replying.
 I&#x27;ve implemented a system that discourages this — repeated
 ghosting eventually results in being removed from the app.- Safety at its core: photo scanning, real-time moderation for
 harassment and grooming patterns, behavioural flagging.- No swiping feed: you spend your time actually talking to
 matches, not browsing profiles.Stack: Supabase, Railway, React Native, and Claude Code did a
lot of the heavy lifting on the backend logic.I&#x27;m a solo builder, this is my first real launch, and I&#x27;d
genuinely appreciate any feedback. Happy to answer anything in
the comments.

Show HN: A personality first matchmaking app

# This is a Expo Framework

Expo is a framework and a platform for universal React Applications. 
It is a set of tools and services built around React Native and native platforms that help you develop, build, deploy,
and quickly iterate on iOS, Android, and web Apps from the same JavaScript/TypeScript codebase.

# Requirements

- Node.js LTS release or greater
- Git
- Watchman for macOS users

# Recommended Tools

- VSCode Editor
- Yarn
- Windows users: PowerShell or Bash via WS

# Installing Expo CLI

```bash
 # install the command line tools
 npm install --global expo-cli
```

## Getting Started 

Create a new app:
At this point we should have Expo CLI installed on our development machine and the Expo Go App on an iOS or Android physical device or emulator.
If not, go back to the Installation guide before proceeding.

# Initializing the project

```bash
# Create a project named my-app.
# Select the "blank" templata when promped
expo init my-app

# Navigate to the project directory
cd my-app
```

First, run the development server:

```bash
expo start
# or
npm start
```

... More content comming soon ...

ReactNative News

News From: hn.algolia.com