Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/TrinaxCode/TrinaxAI/llms.txt

Use this file to discover all available pages before exploring further.

TrinaxAI is an open-source AI assistant that keeps your data completely local. It combines a beautiful cross-platform PWA chat interface, a developer CLI, RAG-powered code search with citations, voice conversations, and image analysis — all powered by Ollama models running on your own hardware.

Quickstart

Get TrinaxAI running on your machine in minutes with a single command.

Installation

Platform-specific guides for Linux, macOS, and Windows.

CLI Reference

Use trinaxai ask, chat, index, research, and more from your terminal.

API Reference

Integrate with TrinaxAI’s FastAPI backend from any client.

What is TrinaxAI?

TrinaxAI is a three-tier local stack: a React PWA on port 3334, a FastAPI RAG engine on port 3333, and Ollama running your models on port 11434. Everything communicates over localhost or your private LAN — no data ever leaves your network.

RAG & Indexing

Hybrid vector + BM25 retrieval with AST-aware code chunking for 15+ languages.

Chat & Models

Auto-routing selects the best model per query — general, code, deep, or fast.

Voice & Vision

Speech-to-text, text-to-speech, and local image analysis with qwen2.5-vl.

Memory & Collections

Persistent facts and separate RAG namespaces for each project.

PWA

Install as a native app on iOS, Android, and desktop with offline support.

Security

Local-only by default. System endpoints require authorization.

Get started in 60 seconds

1

Install on Linux or macOS

curl -fsSL https://raw.githubusercontent.com/TrinaxCode/TrinaxAI/main/install.sh | bash
On Windows, see the Windows installation guide.
2

Open the PWA

Navigate to https://localhost:3334 in your browser. Accept the self-signed certificate prompt and start chatting immediately.
3

Index your code

trinaxai index ~/my-project
TrinaxAI will AST-chunk and embed your code. Ask questions with full source citations.
4

Run a health check

trinaxai doctor
Confirms Ollama, the RAG API, and your models are all running correctly.
TrinaxAI auto-detects your available RAM and selects an appropriate hardware profile (8 GB, 16 GB, max, or ultra). You can override this with --profile during install or TRINAXAI_PROFILE in your .env.

Build docs developers (and LLMs) love