Frequently asked questions

You get the complete VectorBT PRO package: the private GitHub repository, documentation and API references, tutorials, cookbook recipes, release notes, product updates, Discord support, and access to the private community.

We use open-source VectorBT as the trial. It is free, gives you a real feel for the array-based workflow, and lets you decide whether this way of working clicks with you. The public documentation and feature pages show what PRO adds on top.

Pick monthly if you want maximum flexibility, 12-month access if you are ready to commit for a year and save on the monthly price, or lifetime if you prefer one payment with no renewals. The product and member benefits are the same on every active plan.

Pay once and get lifetime access to the complete VectorBT PRO membership with no renewals. This includes the repository and its updates, the private website, Discord, and support, subject to the Terms of Use.

Yes. We credit half of your previous monthly payments toward lifetime access, with a minimum upgrade payment of $150. If you are still in your first month, the full first payment counts as credit. Email olegpolakow@vectorbt.pro and we will calculate the upgrade for you

Yes, organization access is available for internal, non-commercial work. If you want to use VectorBT PRO in a commercial product or service, email olegpolakow@vectorbt.pro and we will find the right setup. The software license has the full details.

GitHub Sponsors is the fastest route: repository access is granted immediately. Other payment routes are processed within 24 hours. For Discord, join the server and post your GitHub username in the verification channel; we will review it and grant access within 24 hours.

Ask us anything about VectorBT PRO on Discord. We help with installation, APIs, research workflows, debugging, and your own implementations. The community is also a great place to compare approaches and learn how other members solve similar problems.

Absolutely. Cancel through your payment provider whenever you like, and your access will stay active until the end of the paid term. Refunds and statutory withdrawal rights depend on applicable law and any additional policy offered by the payment provider. The Terms of Use cover the details.

The latest version you installed keeps working. Access to the private repository, website, Discord, new releases, bug fixes, and support ends with the paid term. You can find more detail in the license remarks.

VectorBT PRO is a code-first toolkit for quantitative analysis, backtesting, and algorithmic trading research. It provides a Python package and a native Rust crate, both designed to run in your own environment and integrate into existing workflows.

VBT is market-agnostic. Members use it with equities, cryptocurrencies, futures, foreign exchange, options, and other instruments. Supply the data you need and model the details that matter for that market, such as contract rules, fees, slippage, and execution behavior.

Yes. Bring Pandas or array-based data, load local files and databases, use a built-in adapter, or write a custom data class. VBT keeps the research layer flexible, so you can connect whichever data source, broker, or exchange fits your workflow.

Yes, as the research and signal engine inside your own trading system. A typical workflow pulls fresh data, runs VBT, extracts the latest signals or orders, and passes them to your broker or exchange interface. Your execution layer handles connectivity, state, and live risk controls. With the MCP server and modern coding agents, building that interface is more approachable than ever.

VBT includes flexible cross-validation, robustness tests, and tools for analyzing out-of-sample behavior and parameter stability. Together, they help you understand whether an idea holds up across different market periods and configurations instead of trusting a single strong backtest.

Yes. The MCP server, CLI commands, LLM-ready documentation, and agent tools help compatible coding agents navigate the APIs, run real VBT code, and automate research workflows. You spend less time looking up syntax and more time working on the idea.

Basic Python and some familiarity with Pandas or NumPy are enough to get started. VBT is code-first, but you do not need to be an expert. Tutorials, cookbook recipes, documentation, Discord, and coding-agent tools can take you from a small example to a serious research workflow.

VBT runs on Linux, macOS, and Windows. We recommend WSL on Windows for the smoothest development experience. You do not need a GPU; CPU and memory needs depend on the size of your data and parameter space. Chunking, caching, parallel execution, and distributed execution help when a workload grows.

Once GitHub access is active, install directly from the private repository with uv or pip. You can also clone it for development, use the included Docker environment, or run it in Google Colab. The private installation guide walks through every option.

Yes. Use VBT in scripts, applications, IDEs, Jupyter notebooks, or Google Colab like any other Python package. The repository also includes a ready-to-build JupyterLab Docker environment.

No. The complete high-level workflow is available from Python, with Numba handling compatible numerical operations. The optional Rust extension comes as platform wheels, so you can use it without setting up a Rust toolchain. The Rust implementation is also available as a native crate if that is where you want to work.

Yes. The core array-based ideas will feel familiar, but PRO is a substantial redesign rather than a drop-in replacement. Start in a dedicated environment, import vectorbtpro instead of vectorbt, and use the current documentation and examples while adapting your code.

New releases arrive through the private GitHub repository. Re-run your uv or pip install command to update, pull the latest changes if you work from a clone, or rebuild the Docker image if that is your setup.

Yes. Once installed, core calculations, local-data workflows, and offline knowledge tools work without an internet connection. Naturally, installing updates, fetching remote data, calling cloud services, and opening the website or Discord still need a connection.