A next-generation framework for building broadcasting apps across every screen — from phones to set-top boxes to smart TVs. Powered by Flutter, with AI-driven developer tooling that accelerates your entire workflow.
Learn MoreFlutterTV is a unified platform that lets developers create beautiful, high-performance broadcasting applications for phones, set-top boxes, smart TVs, and beyond — all from a single Dart codebase.
But this isn't just another cross-platform toolkit. It's built from the ground up with AI tooling for developers — LLM gateways, an agent orchestration layer, a skill/tool registry, and prompt management — so teams ship faster, reuse more, and stop rebuilding the same thing for every platform.
The result: over 90% shared code across targets, a dramatically shorter time-to-market, and a framework that grows with your device fleet — not against it.
Not AI-as-a-feature — AI as a force multiplier. Built-in LLM gateway (DeepSeek, GPT, Claude, on-device models), agent orchestration, skill/tool registry, context management, and prompt engine. Developers wire up intelligent behaviour without starting from scratch.
A single shared Dart core with over 90% code reuse across phones, set-top boxes, and smart TVs of every kind. Platform-specific optimizations live at the embedder level — native performance where it counts, shared logic everywhere else.
Layouts that adapt to the device, not the other way around. TV-optimized grid navigation on the big screen, touch-friendly lists on mobile. Per-platform theming, responsive breakpoints, and accessibility adjustments — all driven by configuration, not hardcoded rules.
Every external concern — players, CMS, analytics, new components, even new target platforms — is a plugin. The core framework stays lean. Plugins register through a central registry with typed contracts and dependency resolution. Add Hisense or tvOS support without touching framework internals.
UI structure, content layout, and theming are configurable at runtime via a CMS abstraction layer. Personalization rules — who sees what, when — are driven by configuration, not code deploys. Update the homepage layout for a new campaign without rolling a build.
TV devices are resource-constrained. The framework is designed for low-latency rendering, lazy component loading, on-device model support, and background prefetching. Every millisecond of startup time is accounted for.
Our initial Proof of Concept targets three key platforms — with an architecture designed to expand across the full landscape of connected-TV devices.
| Platform | Devices | Status |
|---|---|---|
| Android Mobile | Phones, tablets | POC |
| Android TV | Set-top boxes, Fire TV, Chromecast, NVIDIA Shield, operator STBs | POC |
| LG webOS | LG Smart TVs (all recent models) | POC |
| Coming Next | ||
| Samsung Tizen | Samsung Smart TVs, monitors | Planned |
| Apple tvOS | Apple TV HD, Apple TV 4K | Planned |
| Roku | Roku streaming players, Roku TV | Planned |
| Gaming Consoles | Xbox, PlayStation | Planned |
The plugin architecture means every new platform is additive — no rewrites, no forks. The same Dart business logic, a new embedder.
Building broadcasting apps today means maintaining separate codebases for each platform — Android (Java/Kotlin), webOS (JavaScript), tvOS (Swift), Tizen (C++/JS), and more. Each with its own UI toolkit, its own player integration, its own analytics pipeline, its own release cadence.
A single feature — a new content row on the home screen — can take three sprints to land everywhere. Platform quirks multiply regression risk. Delivery timelines stretch because there's no shared foundation.
FlutterTV gives teams a single codebase with a single delivery pipeline. Business logic lives in shared Dart. UI adapts per platform through configuration, not duplication. AI tooling — LLM gateways, agent orchestration, context management — is baked into the framework so developers can wire up intelligent behaviour on day one, not month six.
Less code. Fewer bugs. Faster delivery. That's the point.
Imagine spinning up a new broadcasting app project and having a working three-platform build — Android Mobile, Android TV, LG webOS — within the first hour. A scaffolded project with navigation, theming, player abstraction, and AI plumbing already wired in.
Imagine writing your content recommendation logic once — in Dart — and having it run identically on a $30 streaming stick, a flagship phone, and a premium OLED TV. No per-platform rewrites. No lowest-common-denominator UX.
That's what FlutterTV is building. A framework where business logic is written once, where AI tooling isn't a separate project but the foundation, and where expanding to a new platform means writing an embedder — not rewriting your app.
We're building in phases, validating the architecture at each step.
Core framework, plugin system, theming engine, and context management.
LLM gateway abstraction, agent orchestration, skill/tool registry, prompt management, and on-device model support — all exposed through clean APIs. Define skills once, register them, and any agent in the framework can call them — no per-platform wiring, no model lock-in.
Embedders for Android Mobile, Android TV, and LG webOS with native player and input integration.
Adaptive component library with TV-optimized navigation and responsive grids.
CMS-driven layouts, configuration-based personalization, player abstraction, and analytics pipeline.
Rendering benchmarks, startup-time optimization, memory profiling on constrained devices, and comprehensive test coverage.