--- title: "Real-Device Testing" description: "Test on physical iOS and Android hardware with agent-driven runs, where gestures, performance, and native integrations actually behave." url: https://www.callstack.com/services/real-device-testing updated: 2026-11-09T15:59:44.782Z --- # Real-Device Testing Test on physical iOS and Android hardware with agent-driven runs, where gestures, performance, and native integrations actually behave. _[Agentic Delivery](/agentic-delivery) › [AI-Driven QA & Testing](/capabilities/ai-driven-qa-testing)_ **Test where your users are: on hardware** Simulators lie about performance, gestures, and native integrations. We run agent-driven testing on physical devices through Agent Device, our open-source automation platform - locally or from any CI runner. ## AI-Driven QA & Testing - Industry runs on our tooling · 50M+ OSS downloads / month - 100+ Enterprise projects delivered - React Foundation Founding Member · 300+ commits to React Native · Contributing since 2016 - Expensify uses agent-device for bug evidence. Its usage story covers collecting evidence from mobile apps and profiling. - ~6x Faster prototyping · 60+ Components mapped with full coverage · Figma · Claude · Server · Tailwind · Tokens · React · Cursor · ESLint - 3.7× Lower task cost vs Opus 5.5 - Testing Library · Testing utilities in our open-source toolkit for checking React Native behavior. ## What you get: Real hardware, wired into delivery. - **Physical device execution** Flows run on real iOS and Android hardware, not emulator approximations. - **Built on Agent Device** Our open-source, agent-agnostic platform: auditable, extendable, no vendor lock-in. - **Performance on device tiers** Startup, frame drops, and memory measured on the low-end hardware users actually own. - **Native integration coverage** Camera, biometrics, push, and deep links - the things simulators fake. - **On-prem or cloud devices** Your device lab or hosted hardware, matched to your security policy. - **CI-native runs** Real-device checks triggered from PRs and pipelines like any other job. ## Scope → Build → Scale: Design QA that keeps delivery moving. ### Scope (1 WEEK) We select a product workflow, define where agents act, and agree the controls and success measures. - Workflow and owners defined - Baseline and controls agreed - Token budgets confirmed ### Build (1-12 MONTHS) We connect context, agents, and verification to your tools, then run the workflow on your product backlog. - Running on your product backlog - Speed and quality measured - Business impact validated ### Scale (AS NEEDED) We document the workflow, transfer ownership, and extend it across teams and products based on results. - Your team owns the workflow - Expand based on results - Full knowledge handover ## What shipping at AI speed looks like. - [Building an Agentic SDLC for a Web Launch in Under One Month](/case-studies/building-an-agentic-sdlc-for-a-web-launch-in-under-one-month): 3x estimated delivery acceleration - [Engineering AI Loops for a Cross-Platform Design System](/case-studies/engineering-ai-loops-for-a-cross-platform-design-system): 60+ components mapped with full Code Connect coverage - [Migrating Questrade Edge Mobile to a Faster, More Scalable Architecture](/case-studies/migrating-questrade-edge-mobile-to-a-faster-more-scalable-architecture): 33% fewer re-renders across most features - [Enhancing Application's Performance for Expensify](/case-studies/enhancing-applications-performance-for-expensify): 5x faster message sending time **Stop trusting simulators.** Real-device coverage that runs from your CI and scales with agents. [Add QA agent](/contact) > Their engineers proved to be a crucial part of our mobile app relaunch project. > > — Brian Aznar, Engineering Director, Major League Soccer We’ll connect you with the right expert within 24 hours. ## Let’s find the right path for your product. **Should we build a device lab or use hosted devices?** Depends on volume and policy. Owned labs win on cost at scale and on data control; hosted wins on elasticity. Hybrid setups are common, and we design both. **How many devices do we need?** Fewer than most teams assume. A small matrix covering your top OS versions and performance tiers catches the large majority of device-specific issues. **How do you deal with flaky hardware?** Health checks, automatic re-provisioning, and quarantine for misbehaving devices - lab reliability is an infrastructure problem, and we treat it as one. **What about iOS provisioning and signing?** We automate certificate and profile management as part of the setup. It is painful exactly once. **What does Agent Device cost?** It is open source. You pay for the engineering to build the infrastructure around it, not for a per-device license. **Does this cover TV and desktop?** Yes. Agent Device automates iOS, tvOS, macOS, Android, and Android TV, so coverage follows wherever your product ships. ## Open source - [Agent Device](https://agent-device.dev/): CLI to control iOS and Android devices for AI agents - [React Native Testing Library](https://github.com/callstack/react-native-testing-library): Simple and complete React Native testing utilities that encourage good testing practices - [Reassure](https://github.com/callstack/reassure): Performance testing companion for React and React Native - [React Native Harness](https://github.com/callstackincubator/react-native-harness): Harness brings Jest-style tests to real native environments so you can test TurboModules like you would test your React Native app: with JavaScript. - [Rozenite](https://github.com/callstackincubator/rozenite): Plug-and-play panels for React Native DevTools. Install, enable, and start debugging in minutes.