download

Two apps. Both free.

Trading Agents Lab is for quick analysis: a full agent debate in about a minute. Trading Agents Lab Pro is for full-depth diligence: the upstream project's complete research pipeline, with live tool calls and multi-round debates, in 8 to 15 minutes. Both are free and open source under AGPL-3.0. The difference is depth, time, and what a run consumes, never price.

Both are desktop applications. They run the AI agents on your machine (with your keys, or against your local Ollama), they store everything locally, and they never phone home.

quick analysis

Trading Agents Lab

A full agent debate on any ticker in about a minute.
Small download, light on your machine.
Gentle on API cost: a typical run stays well under a dollar, market data comes from free yfinance, and a local model via Ollama or LM Studio brings a run to zero cost.

Download for macOS

free · open source · AGPL-3.0 · macOS 12+ · Apple Silicon · 165 MB .dmg

full-depth diligence

Trading Agents Lab Pro

The upstream project's full research pipeline: live tool calls, multi-round debates, deeper reports. A run takes 8 to 15 minutes.
Larger download; it bundles the complete research engine.
Sign in with your ChatGPT plan and full-depth runs come at no per-token cost, within your plan's rate limits. With an API key instead, a run makes dozens of model calls and can cost a few dollars on a frontier model.

Download Pro for macOS

free · open source · AGPL-3.0 · macOS 12+ · Apple Silicon · 183 MB .dmg

Both builds are signed and notarised by RBJ Global, so macOS does not flag them as from an unidentified developer. On first launch macOS asks you to confirm an app downloaded from the internet; click Open and it remembers. Both update themselves after the first install. Not sure which to start with? Take the Lite app first; Pro is a separate install and the two live side by side. What Pro adds →

for educational and research purposes only, not investment advice

Build from source

On an Intel Mac, Linux, or Windows? Build the Lite app from source; it runs the same on all three. Pro builds from source too; its repository README has the steps.

Requires Node 20+ and Python 3.13.

step_1

Clone the repository

git clone https://github.com/RBJGlobal/TradingAgentsLab.git
cd TradingAgentsLab

step_2

Install the desktop dependencies

npm --prefix desktop install

step_3

Set up the Python engine

python3.13 -m venv engine/.venv
engine/.venv/bin/pip install -r engine/requirements.txt

step_4

Run the dev stack

npm --prefix desktop run dev

The Electron window opens, the engine sidecar starts in the background, and the status strip flips to ● Running within a couple of seconds.

step_5

Configure a provider

Open Settings → LLM Providers. Paste a key (OpenAI, Anthropic, OpenRouter, Google Gemini, xAI Grok, or MiniMax), or set the app to use a local Ollama model. Then go to Analyze, type a ticker like NVDA, and run.

Step-by-step walkthrough →

Releases and source

Every build, with release notes and checksums, lives on GitHub releases: Trading Agents Lab and Trading Agents Lab Pro. The full source of both is open under AGPL-3.0, so you can read exactly what runs on your machine before you install it.