NimbleEdge launches DeliteAI: Open-Source, On-Device AI Platform
Varun Khare, Neeraj Poddar
Industry’s first Open-Source, On-Device Agentic AI Platform for Mobile Devices
At NimbleEdge, our mission from Day 1 has been clear: build AI that’s personal, private, and accessible. Today, we’re proud to announce the official launch of DeliteAI - world’s first open-source, fully on-device agentic AI platform built for mobile devices. With DeliteAI, developers can now build, deploy and scale AI-native experiences directly on smartphones, with zero reliance on cloud infrastructure.
This launch marks the rollout of a complete developer stack for on-device AI, built for a future where intelligence runs alongside the user, not in the cloud.
Consumer apps have not evolved with AI or delivered experiences beyond the Chat based interfaces. In order for these apps to truly deliver accessible and personal AI, AI workloads need to move closer to where data is created - your phone, your laptop, your edge device. Cloud-based AI creates friction: it depends on connectivity, increases latency, compromises user privacy and drives up costs.
DeliteAI solves these challenges by bringing LLMs, multimodal models and AI agents directly onto user devices. No cloud roundtrips. No GPU farms on Cloud. Just blazing-fast, privacy-first intelligence running locally - all built on open standards.
With DeliteAI, developers can finally deliver intelligent, personalized and secure AI experiences to billions without trading off control, cost or user trust.
DeliteAI is a full-stack platform purpose-built for on-device agentic AI. At launch, it includes:
DeliteAI SDK – A production-ready SDK with an optimized on-device inference engine and the industry’s first Python runtime for orchestrating agentic workflows on mobile devices using Python scripts. These workflows and models can be dynamically updated on-the-fly via a pluggable SaaS backend without requiring new app releases. You can explore our APIs and sample Python workflow scripts here.
Agent Marketplace – A curated library of prebuilt, plug-and-play AI agents for tasks like summarization, recommendations, speech processing, and more.
NimbleEdge AI Assistant – The first fully open source on-device conversational AI assistant with productivity features, customizable for brand-native experiences.
Unlike traditional AI infrastructure, DeliteAI is built for full customization and developer control. Until now, developers lacked unified tooling, standardized runtimes, and a robust marketplace of ready-to-integrate agents to build sophisticated AI-native experiences on smartphones. Using our SDK, developers can:
Run and leverage open models like Llama, Qwen or Gemini Nano natively and bring your proprietary LLMs or traditional Machine Learning models on Android and iOS devices.
Orchestrate agentic workflows in Python, not in limited DSLs or cloud-bound runtimes - all running on-device.
Optimize performance across diverse mobile hardware without writing low-level code.
We’ve abstracted away the complexities of running AI on-device by providing an interface between AI and mobile applications, so developers can focus on building intelligent apps, not battling hardware or managing AI infra and ML ops.
Even though we have just launched DeliteAI in the open - it’s already battle-tested. At NimbleEdge, we have been building our on-device AI platform for 2+ years and have powered AI/ML experiences across 30M+ devices in production for gaming, e-commerce and productivity apps - and now it’s available in the open source for supercharging AI experiences in your apps.
Our on-device Python runtime and SDK integrate seamlessly with industry-standard runtimes like ONNX and ExecuTorch and abstracts away diverse mobile hardware enabling high-performance AI without data ever leaving the user’s device. This also enables enterprise-grade customization; organizations can bring their own models, brand their own assistants and deliver “delightful” AI experiences while fully controlling data flow, inference logic, and user interaction.
DeliteAI is solving the critical trifecta of performance, privacy, and accessibility — making intelligent apps viable at scale.
DeliteAI is designed to be open, modular and extensible. Our commitment to open-source stems from building and leading communities like OpenMined and Istio.
As of today, developers can :
We’re backed by Neotribe Ventures and advised by AI/ML leaders from OpenAI, Inflection AI and UC Berkeley. This ecosystem is just getting started. NimbleEdge’s open-source, on-device approach is a pivotal shift that unlocks new levels of privacy, performance, and developer innovation.
With the launch of DeliteAI, we’re opening the doors to a future where AI belongs to the developer and stays with the user.
Explore the platform on Github Join our Discord community
Start building - AI on-device is here.
Let’s bring intelligence home.
— Varun & Neeraj
Co-founders, CEO and CTO, NimbleEdge
In our previous blog, we covered how NimbleEdge helps capture event streams
Mobile apps across verticals today offer a staggering variety of choices - thousands of titles on OTT apps, hundreds of restaurants on food delivery apps, and dozens of hotels on travel booking apps. While this abundance should be a source of great convenience for users, the width of assortments itself is driving significant choice paralysis for users, contributing to low conversion and high customer churn for apps.
Discover how NimbleEdge integrates with Google’s Gemini Nano to bring blazing-fast, private, on-device AI experiences. Explore the future of local LLMs and on-device intelligence.