Overview
As AI rapidly becomes an essential part of modern mobile applications, most frameworks have quickly adopted tools that simplify integration with OpenAI, Google, Azure, and similar platforms. However, NativeScript—a framework known for its native performance and developer-friendly approach—had a major gap: there was no easy, standardized, or modern SDK to add AI capabilities into mobile apps.
This fragmentation created friction, slowed down innovation, and made AI adoption in NativeScript unnecessarily difficult.
The Challenge
NativeScript developers were increasingly interested in building:
- In-app AI assistants
- Intelligent customer support bots
- Content generation tools
- Natural-language driven experiences
- AI-enhanced workflows inside native apps
But they lacked a unified solution that offered:
Clean APIs
Simple, intuitive interfaces that don't require deep AI knowledge
Cross-Provider Support
OpenAI, Google Gemini, Azure, and more
Streamed Responses
Real-time, token-by-token streaming for better UX
NativeScript Integration
Full compatibility with ViewModels and UI components
Native Performance
Lightweight design that doesn't compromise speed
Modern DX
React-like developer experience, adapted for NativeScript
Developers had the skills. NativeScript had the power. What was missing was an SDK designed for this ecosystem.
The Solution: ns-ai-kit
I created @marrocode/ns-ai-kit, a lightweight, flexible, and modern AI integration layer for NativeScript—inspired by the simplicity and elegance of the Vercel AI SDK.
The goal was simple: Bring real AI capabilities into NativeScript, without asking developers to switch stacks, rewrite their apps, or adopt new frameworks.
Key Features
Easy-to-use AI Chat Interfaces
Simple APIs for building conversational experiences
AI Provider Support
Currently supports OpenAI and Google Gemini (more providers coming soon)
Streaming Out of the Box
Real-time token-by-token responses for better UX
Event-Based Hooks
Deep customization for advanced use cases
NativeScript Native
Fully compatible with ViewModels & UI components
Minimal Setup
Get started with just npm install
Developer Experience
The SDK takes care of request formatting, real-time streaming, error handling, and provider compatibility—allowing developers to focus on experience, not infrastructure.
npm install @marrocode/ns-ai-kitconst res = await this.gemini.chatCompletion({
messages: [{
role: 'user',
content: 'Explain NativeScript in one paragraph.'
}],
});That's it. No boilerplate, no complex setup, no switching frameworks.
Scalability for Future Use Cases
One of the core design principles of the SDK is scalability. NativeScript apps often start small—an MVP, an internal tool, or a simple feature—and grow into larger systems. ns-ai-kit is built to support this evolution without rewriting or restructuring the codebase.
Start Small
- A basic chatbot
- A small Q&A assistant
- A single AI-powered feature
Scale Up
- Multi-step reasoning flows
- AI-powered data extraction
- Real-time conversational interfaces
- Context-aware assistants
- Generative content pipelines
- Enterprise-grade AI features
All while keeping native performance, thanks to NativeScript's architecture and the SDK's lightweight design.
Who Is This For?
NativeScript Developers
Tired of switching frameworks just to test AI ideas
Startups
Wanting to add intelligence to mobile products quickly
Development Teams
That value native performance but need modern AI tools
Enterprises
Already using NativeScript internally and want AI modernization
Makers & Experimenters
Pushing innovation in cross-platform mobile development
Impact & Recognition
The release of ns-ai-kit filled a long-standing gap in the NativeScript ecosystem. Within days:
Featured by NativeScript
Highlighted by the official NativeScript team, validating its impact within the ecosystem
View Official Tweet →Community Adoption
Early adopters began testing and contributing to the project
Strong Interest
Community expressed enthusiasm for expanding AI capabilities in NativeScript
Developer Experience
Praised for bringing "Vercel-style" DX to the NativeScript world
What This Enables
Most importantly, the SDK empowers teams to:
Stay Within Their Stack
No need to switch frameworks or learn new tools
Build Faster
Prototype and ship AI features without friction
Maintain Performance
Keep native-level speed with minimal overhead
Scale Confidently
Start small and grow into enterprise features
Project Status
Stable release with comprehensive documentation
Open to contributions and feature requests
Technical Highlights
TypeScript First
Built with TypeScript for type safety and excellent IDE support throughout the development experience.
Multi-Provider Support
Currently supports OpenAI and Google Gemini with a unified interface, designed to easily add more providers in the future.
Streaming Architecture
Built-in support for real-time streaming responses, providing smooth, token-by-token output for better user experience.
Lightweight Design
Minimal dependencies and optimized for mobile performance, ensuring fast load times and efficient resource usage.