Trading frameworks

Run the open-source trading frameworks you already use as managed cloud workloads. Deploy from GitHub, run in broker-proximate regions, controlled from your AI IDE.

TradingHost does not replace these frameworks. It provides the runtime container, deployment workflow, logs, health checks and region selection around them. Whichever framework matches your strategy, the runtime model stays the same.

Frameworks

Pick the framework you build in

Each framework has its own landing page covering deployment shape, supported venues and the parts of the platform it leans on.

Exchange ConnectivityTalk to every venue from one unified API surface.

Unified exchange API

CCXT

The de-facto unified API for crypto exchanges: 100+ venues through one interface, in six languages, with REST and WebSocket both included. Deploy a CCXT-based strategy from GitHub and run it as an always-on container.

Open CCXT

Strategy BuildersAuthor full trading strategies in Python.

Crypto strategy framework

Freqtrade

Open-source Python crypto framework with one strategy file and five modes: backtest, hyperopt, dry-run, FreqAI and live. Uses CCXT underneath for exchange connectivity, and deploys from GitHub as an always-on container.

Open Freqtrade

Python algo framework

Jesse

MIT-licensed Python algo framework with 300+ built-in indicators, multi-symbol and multi-timeframe backtesting, and no look-ahead bias. Live in a managed container with exchange secrets handled outside the repo.

Open Jesse

Market Making and LiquiditySpread capture, inventory control and venue-aware quoting.

Open-source market making

Hummingbot

Market-making framework with its own native connectors — no CCXT dependency — across 30+ CLOB CEX, CLOB DEX and AMM DEX venues. Run quoting and cross-exchange strategies as long-lived runtimes in venue-proximate regions.

Open Hummingbot

Research and BacktestingBuild the strategy locally. Deploy the same code to production.

Python backtesting

Backtrader

Python backtesting library with 122 built-in indicators and live hooks for Interactive Brokers, Oanda and Visual Chart. The original repo has been quiet since 2023 and a community fork maintains it; TradingHost runs whichever version your repo pins.

Open Backtrader

Vectorised backtesting

vectorbt

Vectorised NumPy backtesting that evaluates thousands of parameter combinations in one pass. Research and backtesting only — the open-source build includes no live trading — so it runs here as a compute-sized research runtime.

Open vectorbt

Trading EnginesHigh-performance execution engines for low-latency strategies.

High-performance engine

NautilusTrader

Rust core with a Python API. Event-driven, deterministic and built for low-latency strategies. Runs as a long-lived container next to the venues that matter.

Open NautilusTrader

Bring your framework, we run it

TradingHost runs whichever framework matches your strategy. CCXT, Freqtrade, Jesse, Hummingbot, Backtrader, vectorbt or NautilusTrader. The runtime, deployment, observability and operational surface are identical across them.

Your repo owns the strategy. TradingHost owns the production environment around it.