QuanuX

The open standard for quantitative trading systems.

QuanuX enables rapid strategy iteration in Python, high-performance execution via C/C++ components, and native connectivity to multiple brokerages.

Run QuanuX locally or on a remote server, enjoy Anaconda-Navigator integration alongside your favorite Data Science, AI and Machine Learning tools, use the full GUI's, code your own, or control headless with our CLI tool quanuxctl.

Open source. Self-hosted. Your data.

What QuanuX Is

QuanuX exists to remove artificial complexity from quantitative trading systems while preserving real complexity where it actually matters.

It provides a clear architectural spine — schemas, tools, strategies, and bridges — and leaves control in the hands of the user.

QuanuX is a platform for building your own systems, your own way.

Core Principles

Modular by Design

Trading systems in QuanuX are composed of independent, reusable components rather than monolithic scripts. Signals, sizing, risk, execution, and configuration live where they belong — and can be reused without rewriting redundant code.

Python-First Iteration

Strategies are developed and iterated in Python to maximize speed, clarity, and experimentation. Where performance matters, QuanuX supports fast C/C++ indicators exposed through Python wrappers.

Explicit Authority

QuanuX enforces explicit Schema Authority and Tool Authority to prevent silent drift. Contracts are defined once and verified continuously. Tooling is registry-defined, inspectable, and constrained by design.

Choice Without Lock-In

QuanuX provides integrations and examples — not mandates. You may use or replace any part of the system. Guardrails exist to guide you, not to trap you.

Architecture
at a Glance

QuanuX separates concerns intentionally.

Clients

Presentation layers (Desktop, Web, Mobile, TV)

The Server

Hosts all business logic and execution control

Strategies

Modular, testable systems

Bridges

Connect to brokerages and data providers

Schemas

Define canonical contracts

Tools

Expose controlled automation through MCP

This separation allows QuanuX to scale from a single laptop to distributed clusters — without changing how strategies are written.

Clients & Platforms

QuanuX clients are purpose-built for their roles.

Desktop and Web clients are built with React, Tailwind, and a lightweight Rust-based Tauri container. These are the primary environments for development, configuration, and deep system interaction.

Mobile, tablet, TV, and other device clients are built with React Native. These targets are intended primarily for monitoring and interaction, not for authoring strategies.

These choices reflect intended use — not enforced limits.

You are free to re-implement any client using any technology if your needs differ.

AI, On Your Terms

QuanuX is AI-ready, but not AI-dependent.

AI tools can be integrated to assist with problem solving, edge-case reasoning, platform extension, and strategy exploration. QuanuX provides agent instruction files that guide AI systems in understanding how the platform is structured and where logic belongs.

You bring your own AI provider — including private or self-hosted models.

  • QuanuX does not proxy prompts.
  • QuanuX does not collect AI data.
  • QuanuX does not centralize intelligence.

AI is a tool. You remain responsible.

Privacy & Ownership

QuanuX is installed on hardware you control.

No telemetry
No tracking
No analytics
No hidden data collection

Your strategies, data, models, and experiments remain yours.
Nothing leaves your system unless you explicitly configure it to.

This is not a feature. It is a design constraint.

Who QuanuX Is For

QuanuX is built for people willing to understand their tools.

  • Independent traders and prop traders
  • Quantitative researchers
  • Funds and institutions that require inspectable systems
  • Students who want real systems, not toy abstractions

You do not need a PhD to use QuanuX.

You do need curiosity, patience, and responsibility.

How to Learn QuanuX

Dossier is a living, curated encyclopedia of architecture, patterns, experiments, and lessons learned. It is editorial and exploratory by design.

Reference is the canonical technical source. It is precise, stable, and machine-consumable.

Dossier explores.

Reference defines.

Supporting QuanuX

QuanuX is free and open source.

If you find it valuable and want to help ensure its independence and longevity, you can support the project on GitHub.

Support the Project

Sponsorship funds ongoing development, infrastructure costs, and documentation.

It is a contribution to the open ecosystem, not a purchase of access.

Pre-Release Notice

QuanuX is currently in v0.0.1 pre-release.

Interfaces, APIs, and documentation may evolve as the platform is refined.

The architectural spine is stable.The system is actively being built.

QuanuX is not here to sell you access.

QuanuX puts you in control.