Pricing

The engine is free. You pay when the work has to leave your laptop.

Everything that computes a number is open source and always will be. What costs money is that the work persists, reaches the person who has to act on it, and adds up into something your investors can be shown.

Open source

It computes
Freeforever

The whole engine, on your machine, offline. No account, no key, no network call on import.

  • Every statistic, test and correction
  • Parameter sweeps, with the trial count derived
  • Study files on your own disk
  • Runs with the wifi off
Read the docs

Solo

It persists
$499per month

One person, one seat, one job — a quant OR a PM. Your runs, studies and decisions stop living in a notebook.

  • Everything in Open source
  • Runs and studies persist, captured as you work
  • Theses, decision journal, monitors
  • Your own track record and calibration
Get access

Pod

It crosses to someone else
$2,499per month

The moment a quant AND a PM are both involved, there is a handoff — and a handoff is a pod. Several sleeves, one record.

  • Everything in Solo
  • Quant to PM handoff, and the surface to judge it
  • Assignments, shared universe, data connectors
  • Ranked signal files out to your own systems
Get access

Fund

It adds up over time
$7,999per month

For a CIO over more than one sleeve, and the committee behind them. Where the periods and the evidence live.

  • Everything in Pod
  • Sleeves, capital allocation, risk budgets
  • The close: review packet, letter, attribution
  • Committee packet, countersignature, diligence export
Get access

Billed monthly, cancel any time. Two months free on an annual term. Access is provisioned by us today, so a paid plan starts with a short conversation rather than a card form.

What we will not do

We never charge for rigour.

There is one common way to ruin an open core, and it is to put the features a person needs to do good work behind the paywall. Every statistic, every correction for how many things you tried, every control that lets the system tell you your idea does not hold up, is in the free tier permanently.

Never gated

The maths, the tests, the trial counting, and every honest degradation. If the system knows a number is weak, it says so at every tier.

Gated on people

Seats, handoffs, and anything that involves a second person reading your work.

Gated on time

Periods, closes, packets and retention. The things that only mean anything once months have passed.

In detail

What is in each plan.

FeatureOpen sourceSoloPodFund
The engine
Every statistic, test and correction
Parameter sweep, trial count derived
Study file, portable and versioned
Offline, no account required
Your data is never stored
It persists
Runs captured as you work
Studies, kept and searchable
Theses and decision journal
Monitors on what would change your mind
Track record and conviction calibration
It crosses to someone else
Seats11 (quant OR PM)Up to 6, mixed typesUnlimited
Sleeves1SeveralSeveral, across pods
Quant to PM handoff and evaluation
Assignments and shared universe
Data connectors
Ranked signal file export
It adds up over time
Sleeves and capital allocation
Risk budgets across pods
Period close and review packet
Committee packet and countersignature
Diligence export, retention and audit
Single sign-on
Support
ChannelCommunityEmailEmailNamed contact

The full stored and never-stored line lives on the Trust page. Computational tooling, not investment advice.

Getting started

Install it first.

The honest order is that your quant runs the engine on your own data and forms a view, and then you pick a plan. Nothing on the free tier expires and none of it is a trial.

Start here
$ pip install alphaengine