flutter_edge_ai
On-device AI for Flutter — Gemma and other LLMs, multimodal, function calling, embeddings, RAG and speech on Android, iOS, Web, macOS, Windows and Linux.
What is flutter_edge_ai?
On-device AI for Flutter — Gemma and other LLMs, multimodal, function calling, embeddings, RAG and speech on Android, iOS, Web, macOS, Windows and Linux.
Overview of flutter_edge_ai
What it is
Flutter Edge AI is an on-device AI SDK for Flutter that runs Gemma and other large language models, multimodal features, function calling, embeddings, retrieval-augmented generation (RAG), and speech on Android, iOS, Web, macOS, Windows, and Linux. It is described as having no server and no cloud. The core package installs models and hands them to separate engine packages; each runtime, store, and feature is a separate opt-in package so an app ships only what it uses.
Key features
- Runs on Android, iOS, Web, macOS, Windows, and Linux.
- Supports multiple engines: LiteRT-LM (
.litertlm), MediaPipe (.task), llama.cpp (.gguf), ONNX Runtime, and built-in OS models (Gemini Nano on Android, Apple Foundation Models on iOS/macOS, Windows AI Foundry, browser Prompt API). - Provides embeddings, on-device RAG with SQLite (
sqlite-vec) or Qdrant Edge vector stores. - Offers speech-to-text, text-to-speech, and voice session loop.
- Includes agent skills (text, JS, native intents, MCP tools).
- Has a Genkit Dart plugin and hybrid routing for on-device and cloud models.
Use cases
- Running large language models locally on mobile and desktop apps without a server.
- Building on-device AI assistants with function calling and RAG.
- Implementing speech interfaces (speech-to-text, text-to-speech).
- Embedding AI into Flutter apps across multiple platforms with a modular, opt-in architecture.
Getting started
See the repository README for installation and quick-start commands.
License and pricing
Licensed under MIT.
Based on the project's public README and repository metadata.
