hummingbot: Hummingbot – open‑source framework for building and running automated crypto‑trading bots
Hummingbot lets you create, configure, and run algorithmic trading bots that work on dozens of centralized and decentralized crypto exchanges. It provides a command‑line tool (hbot) and Docker images for easy installation, a library of ready‑made strategies (market‑making, arbitrage, grid, etc.), and a plug‑in system for exchange connectors. A separate Condor project adds optional LLM‑driven decision making, but the core product is a trading‑automation platform, not a machine‑learning library.
Key points from the README
- Purpose – democratise high‑frequency/algorithmic trading by offering a free, community‑driven bot framework.
- Installation – supports Conda‑based source install, Docker Compose, and a
hbotCLI for scripting bots. - Strategies – scripts, controllers, executors, and legacy V1 strategies covering market‑making, arbitrage, grid, TWAP, etc.
- Exchange connectors – unified APIs for many CEXs (e.g., Binance, KuCoin) and DEXs (e.g., Uniswap, Curve), with both custodial and non‑custodial options.
- AI integration – the optional Condor harness connects LLM‑powered decision logic to Hummingbot via a Telegram/web dashboard.
- Community – active Discord, docs site, newsletters, and a governance model for adding new connectors.
- License – Apache 2.0, fully open source.
Who might use it?
- Crypto traders who want to automate strategies without writing low‑level exchange code.
- Developers building custom bots or extending existing strategies.
- Researchers experimenting with AI‑augmented trading via the Condor harness.
What it is not
- A machine‑learning framework or LLM model library. Its AI component is an optional wrapper; the core product is a trading‑automation engine.
TL;DR – Hummingbot is an open‑source platform for creating and running crypto‑trading bots across many exchanges, with optional LLM‑driven agents via the separate Condor project.
Off‑topic flag: false (the repository is a genuine software project, though its primary focus is automated trading rather than core AI/ML research.)