Callstack Engineers

React Native migrations can go from months to days with AI. Share logic across iOS and Android, reduce duplicated work, and move faster with one codebase. We’ll show you how we use agents in real migration projects to speed up the process and secure outcomes. Join us on April 2, 5 PM 👇 clstk.com/4lFxkuB

Can an app tell you when to push, and when to stop? @tejask built a React Native app around HRV (heart rate variability), that turns wearable data into a simple readiness signal. Watch his talk from React Universe Meetup x @Zalando now ⬇️

Agent Device: AI agents meet mobile automation x.com/i/broadcasts/1…

Early bird tickets for Agent Conf are almost sold out! Last chance to grab them at the lower price. If you’re working with AI agents—or planning to—this is where the real conversations will happen. 📅Warsaw, Sept 17–18. Get your ticket 👇 clstk.com/4rZRyRz

What if your cloud agents could use iOS and Android devices just like that? Introducing Agent Device Cloud: hosted on your dusted Mac Mini, used through 𝚊𝚐𝚎𝚗𝚝-𝚍𝚎𝚟𝚒𝚌𝚎

If it looks like a duck, and quacks like a duck, then it’s probably a 🦆 I guess there’s a new kid on the block 🥸

@grabbou harness has been such a huge help for shipping, i can literally improve Nitro on my phone via Codex or handwritten via GitHub - if CI passes, we're good to co

Introducing Rozenite for Agents 🚀 The agent-facing interface for interacting with React Native DevTools and Rozenite plugins. It enables coding agents to inspect React Native apps - providing access to logs, network activity, React profiling data, and powerful plugin tooling.

"Which Bielik model should I run?" The answer depends on: 🔹your hardware 🔹your latency target 🔹your cost constraints Here’s how quantization levels (FP16, INT8, Q4) map to real deployment environments. Read the article by @LeSiOO ⬇️ clstk.com/4byI21s

In the latest version of agent-device: diff screenshot. Compare differences in UI at a glance.

We will be in Miami doing a deep dive into running generative models on mobile NPUs. If you're deploying AI on devices at the edge, this one will be for you! ⬇️

Presenting native screens from React Native in a brownfield app sounds simple… until you actually try to implement it 😵‍💫 That’s why we built @callstack/brownfield-navigation - a clean, typed way to present native screens from RN. Built for better DX across RN + native teams 🚀

On-device AI can generate structured outputs, call tools, and interact with APIs. That means real production AI in React Native, not just chat. See how models decide when to call tools and return typed data your app can rely on. Watch now ⬇️

React Native Evals: Making AI Code Quality Measurable x.com/i/broadcasts/1…

@p_syche_ @kraenhansen Listen to the full episode of React Universe on Air now🎧 clstk.com/4bkPfC9

If your data and language model live on the device, why should the app need a network round trip to interact with them? @p_syche_ & @kraenhansen talk about: → local-first software → mobile security practices like certificate pinning → voice AI and real engineering stories

agent device goes cloud 🔜👀 testing it with iOS and Android running on remote machines, looks promising

Move LLMs from cloud → device Now you own: • model capabilities & size • memory + disk usage • hardware acceleration support @artus9033 breaks down what running LLMs on-device in React Native really looks like. Watch the talk from React Universe Meetup x @Zalando here ⬇️

gpt-5.4 achieves better results than gpt-5.3 with 3,5x less tokens 🤯

And of course, make sure to star the repo github.com/callstackincub…

Read more on our blog callstack.com/blog/announcin…

Which AI coding model actually writes good React Native code? We couldn't prove it either, so we built a benchmark. React Native Evals: 43 real-world tasks across animations, async state, and navigation. Read more on our blog ⬇️