WorldQuant BRAIN API Client

A lightweight Python wrapper for the WorldQuant BRAIN REST API: authenticate, discover datasets/data fields, and submit alpha expressions for simulation.

⚠️ Important — what this can and can't do

BRAIN does not accept arbitrary external data uploads. There is no endpoint for pushing your own datasets into BRAIN's simulation engine. All alpha expressions are built using the data fields BRAIN itself provides (price/volume, fundamental, alternative datasets it has licensed).

What this client actually does:

• Authenticates with your BRAIN account (handles the biometric/persona verification step BRAIN sometimes requires).
• Lists datasets and data fields available to your account/region/universe.
• Submits FASTEXPR alpha expressions for simulation and polls until results are ready.
• Summarizes key result stats (Sharpe, fitness, turnover, check pass/fail).

What it does not do:

• Upload or inject external/custom data into BRAIN.
• Guarantee "accurate" data — it surfaces exactly what BRAIN's API returns, nothing more. Treat BRAIN's own published data field descriptions as the source of truth for what each field actually measures.

If your goal is to bring outside research or ideas into your alpha construction, the realistic path is: do that research separately, then hand-translate or use an LLM to help translate the idea into a FASTEXPR expression built from BRAIN's existing fields — then simulate it with this client. That's also the pattern used by most existing open-source BRAIN tools that "use AI" alongside BRAIN.

Setup

pip install requests
cp credentials.example.json credentials.json
# edit credentials.json with your real email & password

Never commit credentials.json to version control. Add it to .gitignore if you put this in a git repo.

Usage

from wq_client import WQClient

client = WQClient(credentials_file="credentials.json")
client.login()

# Discover datasets
datasets = client.get_datasets(region="USA", universe="TOP3000", delay=1)

# List fields in a dataset
fields = client.get_data_fields(dataset_id=datasets[0]["id"],
                                 region="USA", universe="TOP3000", delay=1)

# Submit and simulate an alpha expression
result = client.simulate_alpha(
    "rank(ts_mean(close, 5) / ts_mean(close, 20))",
    region="USA", universe="TOP3000", delay=1
)
print(result)

Or just run the included example end-to-end:

python example_usage.py

Files

• wq_client.py — the API client (auth, datasets, fields, simulation)
• example_usage.py — runnable example exercising all major methods
• credentials.example.json — template; copy to credentials.json and fill in

Notes on rate limits & batching

BRAIN limits how many simulations an account can run concurrently. The simulate_batch() method submits expressions sequentially rather than firing them all in parallel, to stay within typical account limits. If you need true concurrency, check your account's specific simulation slot limit on the BRAIN platform first.

brain-world