| Branch |
Version |
Status |
master |
 |
 |
nightly |
 |
 |
develop |
 |
 |
| Platform |
Rust |
Python |
Linux (x86_64) |
1.98.1 |
3.12-3.14 |
Linux (ARM64) |
1.98.1 |
3.12-3.14 |
macOS (ARM64) |
1.98.1 |
3.12-3.14 |
Windows (x86_64) |
1.98.1 |
3.12-3.14 |
Introduction
NautilusTrader is an open-source, production-grade, Rust-native engine for multi-asset,
multi-venue trading systems.
The system spans research, deterministic simulation, and live execution within a single
event-driven architecture, with Python serving as the control plane for strategy logic,
configuration, and orchestration.
This separation provides the performance and safety of a compiled trading engine with
the flexibility of Python for system composition and strategy development.
Trading systems can also be written entirely in Rust for mission-critical workloads.
The same strategy and execution-algorithm code can run across backtest and live systems, reducing
deployment divergence. Live execution still introduces venue, transport, timing, persistence,
external-activity, and reconciliation behavior that a simulation may not reproduce. See
Backtest and live differences.
NautilusTrader is asset-class-agnostic. Any venue with a REST API or WebSocket feed can be
integrated through modular adapters. Current integrations span crypto exchanges (CEX and
DEX), traditional markets (FX, equities, futures, options), and betting exchanges.
Features
- Fast: Rust core with the mimalloc allocator and asynchronous networking using tokio.
- Reliable: Type- and thread-safety backed by Rust, with optional Redis-backed state persistence.
- Portable: Runs on Linux, macOS, and Windows. Deploy using Docker.
- Flexible: Modular adapters integrate any REST API or WebSocket feed.
- Advanced: Time in force
IOC, FOK, GTC, GTD, DAY, AT_THE_OPEN, AT_THE_CLOSE, advanced order types and conditional triggers. Execution instructions post-only, reduce-only, and icebergs. Contingency orders including OCO, OUO, OTO.
- Customizable: User-defined components, or assemble entire systems from scratch using the cache and message bus.
- Backtesting: Multiple venues, instruments, and strategies simultaneously using historical quote tick, trade tick, bar, order book, and custom data with nanosecond resolution.
- Live: Identical strategy implementations between research and live deployment.
- Multi-venue: Run market-making and cross-venue strategies across multiple venues simultaneously.
- AI Training: Engine fast enough to train AI trading agents (RL/ES).
nautilus - from ancient Greek 'sailor' and naus 'ship'.
The nautilus shell consists of modular chambers with a growth factor which approximates a logarithmic spiral.
The idea is that this can be translated to the aesthetics of design and architecture.
Why NautilusTrader?
Trading strategy research is often conducted in Python using vectorized approaches, while
production trading systems are implemented separately using event-driven architectures in
compiled languages.
NautilusTrader removes this separation.
A Rust-native core provides a deterministic event-driven runtime for both research and live
execution, while Python serves as the control plane. The same architecture, execution
semantics, and time model operate across both environments, allowing strategies to move
from research to production without reimplementation.
Python bindings are provided via PyO3 for the Rust-native v2 runtime.
During the v2 transition, v1 receives only critical security backports on the develop_v1 branch.
See the v2 migration guide for migration steps and compatibility details.
No Rust toolchain is required to install prebuilt wheels.
This project makes the Soundness Pledge:
"The intent of this project is to be free of soundness bugs.
The developers will do their best to avoid them, and welcome help in analyzing and fixing them."
[!NOTE]
MSRV: NautilusTrader relies heavily on improvements in the Rust language and compiler.
As a result, the Minimum Supported Rust Version (MSRV) is generally equal to the latest stable release of Rust.
Integrations
NautilusTrader is modularly designed to work with adapters, enabling connectivity to trading venues
and data providers by translating their raw APIs into a unified interface and normalized domain model.
The following integrations are currently supported; see docs/integrations/ for details:
- ID: The default client ID for the integrations adapter clients.
- Type: The type of integration (often the venue type).
For Lighter on Robinhood, LIGHTER_ROBINHOOD is the venue and explicit client ID to register. The
shared Lighter factory keeps LIGHTER as its compatibility default.
Status
planned: Planned for future development.building: Under construction and likely not in a usable state.beta: Completed to a minimally working state and in a beta testing phase.stable: Stabilized feature set and API, the integration has been tested by both developers and users to a reasonable level (some bugs may still remain).
See the Integrations documentation for further details.
Roadmap
The Roadmap outlines NautilusTrader's strategic direction.
Current priorities include stabilizing the Rust-native core, improving documentation, and enhancing code ergonomics.
The open-source project focuses on single-node backtesting and live trading for individual and small-team quantitative traders.
UI dashboards, distributed orchestration, and built-in AI/ML tooling are out of scope to maintain focus on the core engine and ecosystem sustainability.
New integration proposals should start with an RFC issue to discuss suitability before submitting a PR.
See Community-contributed integrations for guidelines.
Security
Security is a priority for the NautilusTrader project, and we value the work of those who help
identify and resolve vulnerabilities. We apply layered controls across the development and release
lifecycle, with signed releases, continuous vulnerability management, and transparent development
practices:
- Source and review controls: CODEOWNERS gate critical infrastructure, dependency manifests, and
lock files; protected branches require signed commits and passing CI; release tags are immutable;
and Rust dependencies are sourced only from crates.io.
- Dependency intake: lock files pin every dependency with cryptographic checksums, third-party
Python packages install from wheels only, new dependency and tooling versions observe a
publication cooldown before adoption, cargo-vet audits Rust provenance, and cargo-deny checks Rust
dependencies against an allow list of licenses compatible with NautilusTrader's
LGPL-3.0-only
license.
- Scanning and fuzzing: Gitleaks secret screening and Zizmor Actions auditing run pre-commit;
CodeQL runs on PRs to
master and pushes to nightly; cargo-audit, cargo-deny, cargo-vet,
OSV Scanner, and pip-audit run on audit-relevant PRs and daily schedules; cargo-fuzz targets cover
selected adapter and signing surfaces.
- Build and release integrity: GitHub Actions are pinned to commit SHAs, CI runners are hardened
with egress allow-listing, Python artifacts carry SLSA build provenance, container images are
Sigstore-signed with attested SPDX SBOMs, and PyPI and crates.io publishing uses OIDC Trusted
Publishing gated to a protected
release environment that never runs pull request or fork code.
- Runtime cryptography: TLS and most runtime cryptography use
aws-lc-rs, the Rust binding for AWS-LC, with Ed25519 signing
via ed25519-dalek.
The OpenSSF Scorecard badge above is one automated repository-health signal; it complements manual
review, CI hardening, and security audits rather than replacing them.
Reporting a vulnerability
Report privately through
GitHub Security Advisories,
or email [email protected] (PGP key available on request). We acknowledge reports within
48 hours and patch critical vulnerabilities within 30 days.
A careful vulnerability report takes real time and effort. We appreciate that, and unless you prefer
to remain anonymous, we credit reporters in the relevant security advisory and release notes.
The Security Policy details scope, coordinated disclosure, and step-by-step release
verification. The Security Architecture
describes the release supply chain end-to-end. For the full policies, see the
Responsible Disclosure and
Supply Chain Security policies; CI/CD security is
documented in .github/OVERVIEW.md.
Versioning and releases
[!WARNING]
NautilusTrader is still under active development. Some features may be incomplete, and while
the API is becoming more stable, breaking changes can occur between releases.
We strive to document these changes in the release notes on a best-effort basis.
We aim to follow a bi-weekly release schedule, though experimental or larger features may cause delays.
Branches
We aim to maintain a stable, passing build across all branches.
master: Reflects the source code for the latest released version; recommended for production use.nightly: Daily snapshots of the develop branch for early testing; merged at 14:00 UTC and as required.develop: Active development branch for contributors and feature work.
[!NOTE]
The v2 release-candidate line is the transition toward a stable API for version 2.x.
Once this milestone is reached, we plan to implement a formal deprecation process for any API changes.
This approach allows us to maintain a rapid development pace for now.
Precision mode
NautilusTrader supports two precision modes for its core value types (Price, Quantity, Money),
which differ in their internal bit-width and maximum decimal precision.
- High-precision: 128-bit integers with up to 16 decimals of precision, and a larger value range.
- Standard-precision: 64-bit integers with up to 9 decimals of precision, and a smaller value range.
[!NOTE]
By default, the official Python wheels ship in high-precision (128-bit) mode on all supported platforms.
For pure Rust crates, high-precision works on all platforms (including Windows) since Rust handles
i128/u128 via software emulation. The default is standard-precision unless you explicitly enable
the high-precision feature flag.
See the Installation Guide for further details.
Rust feature flag: To enable high-precision mode in Rust, add the high-precision feature to your Cargo.toml:
[dependencies]
nautilus_model = { version = "*", features = ["high-precision"] }
Installation
We recommend using the latest supported version of Python and installing nautilus_trader inside a virtual environment to isolate dependencies.
There are two supported ways to install:
- Pre-built binary wheel from PyPI or the Nautech Systems package index.
- Build from source.
[!TIP]
We highly recommend installing using the uv package manager with a "vanilla" CPython.
Conda and other Python distributions may work but aren't officially supported.
From PyPI
This repository and the documentation cover v2. To install
the v2 release-candidate wheels from PyPI using Python's pip package manager:
pip install -U nautilus_trader --pre
The v2 release-candidate wheels use 2.0.0rcN versions and are intended for community testing
before the final 2.0.0 release. We do not recommend using release candidates in production
environments, such as live trading controlling real capital.
The --pre flag is required until 2.0.0 is released. Without it, pip installs the latest stable
v1 wheel, whose Python API differs from the v2 documentation:
# Installs the latest stable v1 wheel, which cannot run the v2 documentation
pip install -U nautilus_trader
Install optional dependencies for interactive tearsheets and charts with the visualization extra:
pip install -U "nautilus_trader[visualization]" --pre
See the Installation Guide for details.
From the Nautech Systems package index
The Nautech Systems package index (packages.nautechsystems.io) complies with PEP-503 and hosts both stable and development binary wheels for nautilus_trader.
This enables users to install either the latest stable release or pre-release versions for testing.
Stable wheels
Stable wheels correspond to official releases of nautilus_trader on PyPI, and use standard versioning.
To install the latest stable release:
pip install -U nautilus_trader --index-url=https://packages.nautechsystems.io/simple
[!TIP]
Use --extra-index-url instead of --index-url if you want pip to fall back to PyPI automatically.
Development wheels
The main package index publishes v2 development wheels from both the nightly and develop
branches, allowing users to test features and fixes ahead of stable releases.
This process also helps preserve compute resources and provides easy access to the exact binaries tested in CI pipelines,
while adhering to PEP-440 versioning standards:
develop wheels use the version suffix .devYYYYMMDD+run.nightly wheels use .devYYYYMMDD when the base version is already a pre-release, and
aYYYYMMDD otherwise.
| Platform |
Develop |
Nightly |
Linux (x86_64) |
✓ |
✓ |
Linux (ARM64) |
- |
✓ |
macOS (ARM64) |
- |
✓ |
Windows (x86_64) |
- |
✓ |
[!WARNING]
We do not recommend using development wheels in production environments, such as live trading controlling real capital.
Installation commands
By default, pip will install the latest stable release. Adding the --pre flag ensures that pre-release versions, including development wheels, are considered.
To install the latest available pre-release (including development wheels):
pip install -U nautilus_trader --pre --index-url=https://packages.nautechsystems.io/simple
Available versions
You can view all available versions of nautilus_trader on the package index.
To programmatically fetch and list available versions:
curl -s https://packages.nautechsystems.io/simple/nautilus-trader/index.html | sed -n 's/.*<a href="\([^"]*\)".*/\1/p' | awk -F'#' '{print $1}' | sort
[!IMPORTANT]
On Linux, confirm your glibc version with ldd --version and ensure it reports 2.35 or newer before installing binary wheels.
Branch updates
develop branch wheels (.devYYYYMMDD+run): Build and publish continuously with every merged commit.nightly branch wheels (.devYYYYMMDD or aYYYYMMDD): Build and publish daily when we
automatically merge the develop branch at 14:00 UTC (if there are changes).
Retention policies
develop branch wheels: We retain only the most recent wheel build.nightly branch wheels: We retain only the 30 most recent publication dates per platform.
Verifying build provenance
All release artifacts published by the project carry cryptographic attestations
generated by the CI/CD pipeline:
- Python wheels and source distribution (PyPI, GitHub Releases, Nautech Systems package index): SLSA build provenance.
- Docker images (
ghcr.io/nautechsystems/nautilus_trader, ghcr.io/nautechsystems/jupyterlab): keyless cosign signatures plus SPDX SBOM attestations.
Both are issued via Sigstore and bound to a specific
commit SHA, so verification ensures the artifact was produced by the official
NautilusTrader GitHub Actions workflow and has not been tampered with since.
For step-by-step verification commands, see Verifying releases in SECURITY.md.
[!IMPORTANT]
Verification requires the GitHub CLI (gh) for Python artifacts
and cosign for Docker images.
Development wheels from develop and nightly branches are also attested.
From source
It's possible to install from source using pip if you first install the build dependencies as specified in the pyproject.toml.
Install rustup (the Rust toolchain installer):
Enable cargo in the current shell:
Linux and macOS:
source $HOME/.cargo/env
Windows:
Install clang (a C language frontend for LLVM):
Linux (also installs lld, used as the Rust linker for faster builds):
sudo apt-get install clang lld
macOS:
xcode-select --install
Windows:
Add Clang to your Build Tools for Visual Studio 2022:
- Start | Visual Studio Installer | Modify | C++ Clang tools for Windows (latest) = checked | Modify
Enable clang in the current shell:
[System.Environment]::SetEnvironmentVariable('path', "C:\Program Files\Microsoft Visual Studio\2022\BuildTools\VC\Tools\Llvm\x64\bin\;" + $env:Path,"User")
Verify (any system):
from a terminal session run: clang --version
Install uv (see the uv installation guide for more details):
Clone the source with git, then sync its dependencies from the project root:
git clone --branch develop --depth 1 https://github.com/nautechsystems/nautilus_trader
cd nautilus_trader
make sync
[!NOTE]
The --depth 1 flag fetches just the latest commit for a faster, lightweight clone.
Use the uv project environment.
The uv project environment lives at python/.venv, beside python/pyproject.toml. Run direct uv
project commands from python/ or pass --project python from the repository root.
For PyO3 compilation on Linux and macOS, run these commands from the repository root after
make sync in Bash or Zsh. For Fish commands, see the developer guide's
environment setup.
Set PYO3_PYTHON in each shell to this checkout's python/.venv/bin/python; replace any saved
export that still points to the root .venv/bin/python.
# Set the Python executable path for PyO3
export PYO3_PYTHON="$PWD/python/.venv/bin/python"
# Linux only: Set the library path for the uv-managed Python runtime
PYTHON_LIB_DIR="$("$PYO3_PYTHON" -c 'import sysconfig; print(sysconfig.get_config_var("LIBDIR"))')"
export LD_LIBRARY_PATH="$PYTHON_LIB_DIR${LD_LIBRARY_PATH:+:$LD_LIBRARY_PATH}"
# Required for Rust tests when using uv-installed Python
export PYTHONHOME="$("$PYO3_PYTHON" -c 'import sys; print(sys.base_prefix)')"
[!NOTE]
The LD_LIBRARY_PATH export is Linux-specific and not needed on macOS.
The PYTHONHOME variable is required when running make cargo-test with a uv-installed Python.
Without it, tests that depend on PyO3 may fail to locate the Python runtime.
Build and install NautilusTrader in release mode:
make build
See the Installation Guide for other options and further details.
Redis
Using Redis with NautilusTrader is optional and only required if configured as the backend for a
cache database or message bus.
See the Redis section of the Installation Guide for further details.
Makefile
A Makefile is provided to automate most installation and build tasks for development. Some of the targets include:
make install: Installs in release build mode with all dependency groups and extras.make install-debug: Same as make install but with debug build mode.make sync: Installs Python dependencies without building the package.make build: Builds and installs the package in release mode (default).make build-debug: Builds and installs the package in debug mode.make build-wheel: Builds a wheel in release mode.make cargo-test: Runs all Rust crate tests using cargo-nextest.make clean: Deletes build artifacts, caches, and build directories.make distclean: CAUTION Removes all artifacts not in the git index when run with
FORCE=1. This includes source files which have not been git added.make docs: Builds the Python documentation with Sphinx and the Rust documentation with Cargo.make pre-commit: Runs the pre-commit checks over all files.make ruff: Runs Ruff over all files using python/pyproject.toml (with autofix).make pytest: Runs all tests with pytest.make cargo-ci-benches: Builds the Rust benchmarks used by CI.make cargo-codspeed-build: Builds the Rust benchmark subset used for CodSpeed comparisons.
[!TIP]
Run make help for documentation on all available make targets.
[!TIP]
See the crates/infrastructure/TESTS.md file for running the infrastructure integration tests.
Examples
Indicators and strategies can be developed in Python or Rust. For performance and
latency-sensitive applications, we recommend Rust. Below are some examples:
Docker
Docker containers are built with the following variant tags:
nautilus_trader:latest has the latest release version installed.nautilus_trader:nightly has the head of the nightly branch installed.jupyterlab:latest has the latest release version installed along with jupyterlab and an
example backtest notebook with accompanying data.jupyterlab:nightly has the head of the nightly branch installed along with jupyterlab and an
example backtest notebook with accompanying data.
You can pull the container images as follows:
docker pull ghcr.io/nautechsystems/<image_variant_tag> --platform linux/amd64
You can launch the backtest example container by running:
docker pull ghcr.io/nautechsystems/jupyterlab:nightly --platform linux/amd64
docker run -p 8888:8888 ghcr.io/nautechsystems/jupyterlab:nightly
Then open your browser at the following address:
http://127.0.0.1:8888/lab
[!WARNING]
Examples use log_level="ERROR" because Nautilus logging exceeds Jupyter's stdout rate limit,
causing notebooks to hang at lower log levels.
Development
We aim to provide the most pleasant developer experience possible for this hybrid Rust and Python codebase.
See the Developer Guide for helpful information.
Nautilus Engineering maintains the
engineering standards, lint configuration, pre-commit definitions, tool pins, and repository
checks shared across Nautilus projects. This repository vendors selected files at the revision
recorded in .nautilus-engineering.lock; project-specific policy and CI wiring remain local.
[!TIP]
Run make build-debug to compile after Rust changes for the most efficient development workflow.
After changes to PyO3 bindings, stub annotations, or wrapped Rust docs, regenerate the
generated Python artifacts from the repository root:
make py-stubs
Commit the generated .pyi files and PyO3 wrapper doc comments changed by this target. See
Generated Python artifacts for details.
Testing with Rust
cargo-nextest is the standard Rust test runner for NautilusTrader.
Its key benefit is isolating each test in its own process, ensuring test reliability
by avoiding interference. See the Rust testing guidance
for details about full-suite support and the limits of plain cargo test.
You can install cargo-nextest by running:
cargo install cargo-nextest
[!TIP]
Run Rust tests with make cargo-test, which uses cargo-nextest with an efficient profile.
Contributing
Thank you for considering a contribution to NautilusTrader. We welcome high-quality work that
improves the project. Before starting a substantial change, open an
issue to discuss the problem and approach
with the team. Small, self-contained fixes do not require prior agreement.
Before getting started, be sure to review the open-source scope outlined in the project's roadmap to understand what's in and out of scope.
Once you're ready to start working on your contribution, make sure to follow the guidelines
outlined in the CONTRIBUTING.md file. This includes signing a Contributor License Agreement (CLA)
to ensure that your contributions can be included in the project.
[!NOTE]
Pull requests should target the develop branch (the default branch). This is where new features and improvements are integrated before release.
Thank you again for your interest in NautilusTrader! We look forward to reviewing your contributions and working with you to improve the project.
Join our community of users and contributors on Discord to chat
and stay up-to-date with the latest announcements and features of NautilusTrader. Whether you're a
developer looking to contribute or just want to learn more about the platform, all are welcome on our Discord server.
[!WARNING]
NautilusTrader does not issue, promote, or endorse any cryptocurrency tokens. Any claims or communications suggesting otherwise are unauthorized and false.
All official updates and communications from NautilusTrader will be shared exclusively through https://nautilustrader.io, our GitHub, our Discord server, or our verified X (Twitter) account: @NautilusTrader.
If you encounter any suspicious activity, please report it to the appropriate platform and contact us at [email protected].
License
The source code for NautilusTrader is available on GitHub under the GNU Lesser General Public License v3.0.
Contributions to the project are welcome and require the completion of a standard Contributor License Agreement (CLA).
NautilusTrader™ is developed and maintained by Nautech Systems, a technology
company specializing in the development of high-performance trading systems.
For more information, visit https://nautilustrader.io.
Use of this software is subject to the Disclaimer.
© 2015-2026 Nautech Systems Pty Ltd. All rights reserved.
一个生产级的算法交易引擎：核心是 Rust 原生的事件驱动内核，由 Python 来驱动。回测和实盘交易共用同一套架构和执行语义，所以策略从研究到上线不需要重写。