MarketlabGet started

Your local tradingworkbench.

Research markets, backtest Python strategies, and run live trading jobs. All from your own machine.

Install Marketlab Read the quickstart
Built for the work behind the trade.Explore Marketlab
Built for your workflow. Ready for your agent.

Your AI agent.
A trading terminal.

Give a terminal-capable agent access to Marketlab. Research markets, write Python strategies, backtest ideas, and prepare execution from the tools you already use.

Explore Marketlab for agents

Codex · Claude Code · Gemini CLI · OpenCode & more

From the first idea.
To the running strategy.

A connected workflow for traders who build. Explore the data, test your logic, and stay close to execution.

research.ipynbJupyterLab
In [1]: mlab.history(...)In [2]: history.source(CANDLES)
ExploreAnalyzeVisualize

Follow your curiosity.

Explore historical market data in Jupyter. Use your Python environment and the libraries you already know.

Explore research
strategy.py
Historical data
Sequential backtest

Put your logic to the test.

Backtest a plain Python file. Inspect its behavior before deciding what belongs in a live market.

Write a strategy
One job. Four clear steps.
01Preview the plan
02Deploy to the daemon
03Inspect orders and fills
04Stop the job

Stay with the trade.

Run a strategy or built-in bot. Follow job state and execution after your deploying terminal closes.

Meet the runtime

Different markets.
One working language.

Connect market data and live execution through one local runtime.

Market data

BinanceSpot & futures
HyperliquidSpot, perps & outcomes
BULKPerpetuals
MarketlabResearch / Backtest / Execute
BULKPerpetuals
HyperliquidSpot, perps & outcomes
HyperLinkSpot & perpetuals

Live execution

Read from one source. Execute on another.
Data access and execution support vary by provider.

Explore supported markets

Your logic.
Or a head start.

Write your own Python strategy, work an order with an execution algorithm, or configure a market-making bot.

01

Work an order

TWAP, VWAP and OIWAP execution algorithms.

02

Make a market

Grid, Mid-Price and Volume-Mid quoting bots.

03

Build your own

Plain Python files. Your environment. Your rules.

A few things
worth knowing.

Do I need trading credentials to get started?

No. You can read supported public market data first. Private account data and live execution require the appropriate provider credentials.

Where does Marketlab run?

The CLI runs on your machine. The daemon can run natively or in Docker, and you can connect to a separate installation over SSH. Supported systems are macOS Apple Silicon, Linux x64, and Linux arm64.

Can I trade directly from a notebook?

Jupyter is the research environment. Sequential backtesting and live deployment use plain Python strategy files. These are separate workflows.

What does a dry run do?

A dry run previews and validates the execution plan without signing or submitting an order. It does not predict fills or guarantee how a live strategy will perform.

What happens when I close the terminal?

The daemon keeps deployed jobs and execution state running. Your host, daemon, network connection, and execution venue still need to remain available.

Your next idea
starts here.

Install Marketlab. Read your first market.
Build from there.

Installation guide
Install the CLI
curl -fsSL https://marketlab.sh/install.sh | bash

macOS Apple Silicon · Linux x64 / arm64