ONNX-Powered Local AI

Client-Side AI
Runs on Your Device

No API calls. No server costs. Complete privacy. EdgeMind runs powerful AI models directly in your browser using WebAssembly and ONNX Runtime.

Powered by Custom ONNX Models

Optimized quantized models that run entirely on-device

🤖

TaskWizer LLM

666M parameters

Context16K tokens
Size~400MB (Q4)
Speed28 tok/sec
🔍

all-MiniLM-L6-v2

Embeddings

Dimension384
Size23MB (Q4)
Speed2,400/sec
🎤

Speech Models

STT + TTS

STT55MB
TTS147MB
FormatONNX Q4

Why EdgeMind?

💰

Zero Server Costs

No per-token billing. No API calls for inference. Run locally.

🔒

Complete Privacy

Data never leaves your device. No server-side processing.

📴

Offline First

Works without internet after initial model download.

Sub-50ms Latency

No network round-trips. Instant responses.

📦

< 2GB Memory

Optimized quantized ONNX models fit in memory.

🌐

Any Platform

Browser, Node.js, React Native, edge runtime.

Get Started

Install the SDK and start building AI-powered applications with just a few lines of code.

1

Install the SDK

npm install @edgemind/js
2

Initialize

const client = new EdgeMind();
3

Start chatting

const { content } = await client.llm.chat({ messages: [...] });
example.ts
import { EdgeMind } from '@edgemind/js'

const client = new EdgeMind();

await client.models.load('taskwizer-llm');

const { content } = await client.llm.chat({
messages: [{ role: 'user', content: 'Hello!' }]
});