Local-first

Chat with local LLM models right inside your browser.

Privacy-first extension for local AI chat with multi-provider support. Connect Ollama, LM Studio, llama.cpp, or OpenAI-compatible servers (vLLM, LocalAI, KoboldCPP) and run conversations directly in your browser with provider-aware routing and local-first workflows.

Install Extension View Source

Local-first Privacy model

∞ AI models

0* Mandatory cloud APIs

*You control the endpoint. If you configure a remote provider, your data will go to that provider.

Works on Chrome, Brave, Edge, Opera, and experimental Firefox.

Features

Powerful, local-first capabilities

Everything you need for productive AI conversations, with full control over provider endpoints.

Multi-Provider Local Models

Use local models through Ollama, LM Studio, llama.cpp, or compatible OpenAI servers with capability-aware routing.

Ollama, LM Studio, llama.cpp, vLLM, LocalAI, KoboldCPP
Provider-aware model selection + capabilities
Streaming responses + stop control

Multi-Chat Sessions

Organize conversations with multiple sessions stored locally in IndexedDB.

Multiple concurrent chats
Session persistence + history
Tab refresh awareness

Enhanced Content Extraction

Modern extraction with lazy-loading support and site-specific strategies for clean context capture.

Defuddle integration for parsing
Smart scroll strategies
GitHub & YouTube extraction helpers

Advanced Text-to-Speech

Speech synthesis with searchable voices, pitch/rate controls, and per-message playback.

Language grouping + search
Custom text preview
Cross-browser voice loading

Prompt Templates

Pre-built prompts for summarization, translation, and code help.

Summarize content
Translate text
Custom templates

File Upload & Processing

Upload PDFs, text, and DOCX files and auto-embed them for semantic retrieval (no OCR pipeline required).

PDF, TXT, MD, DOCX support
Background embedding progress
Configurable max file size

Backup & Restore

Export and restore full data snapshots with manifested ZIP backups.

Full backup/export flows
Restore with partial failure reporting
Versioned manifests

Advanced Configuration

Full control over provider base URLs, model settings, and exclusions.

Provider base URL control
Selected model ref + capability checks
Debug logging

Search & Export

Search UX enhancements with scoped tabs and print/PDF export.

Result grouping + scoped tabs
Print/PDF export entrypoint
Local-only processing

Privacy First

Local-first by default with user-controlled endpoints.

No mandatory cloud APIs
Local data storage
Open source code

Stack

Built with modern technologies

Optimized for performance, reliability, and extensibility.

React

WXT

TypeScript

shadcn/ui

Dexie.js

Readability

Setup

Get started in 3 easy steps

Quick setup path to get chatting with local models.

Step 1

Start a local provider

Run Ollama, LM Studio, or llama.cpp and set the provider URL inside the extension settings.

Step 2

Add the extension

Install from the Chrome Web Store and pin the extension for easy access.

Step 3

Start chatting

Open the side panel, pick a model, and chat on any website. CORS issue? Follow the setup guide.

Read the setup guide

Install Now - It’s Free

Links