Skip to content
AlgoBeam

Documentation

Learn the platform, then automate it

Every surface in AlgoBeam, every function in AlgoBeamScript and every option in the AlgoBeamTS runtime — written as prose you can read start to finish, with search that answers questions instead of matching keywords.

AlgoBeamTS v4.2.031 pages · 6 sectionsUpdated Aug 25, 2026
Popular:
[ 01 ]Contents

The whole manual, in six parts

Read it in order if you are new, or jump straight to the reference you came for. Nothing here is generated — every page is written and kept current by the people who build the thing it documents.

Getting Started

5 pages

The shortest path from an empty editor to a working indicator, a backtest and a result you can read honestly.

Platform

7 pages

Every hosted surface: charts and layouts, screeners, alerts, watchlists, heatmaps, replay and portfolio analytics.

AlgoBeamScript

9 pages

The language reference — file structure, series semantics, the standard library, plotting and strategy orders.

AlgoBeamTS Runtime

5 pages

Installing AlgoBeamTS, writing providers, streaming live bars and putting a script behind an HTTP endpoint.

Data

3 pages

What we cover, how far history reaches, how bars are aggregated, and the rate limits your key runs against.

Account

2 pages

Plans, proration, invoices and refunds, plus seats, roles and shared team workspaces.

[ 02 ]Quickstart

Three minutes to a printed indicator value

AlgoBeamTS is one dependency. Install it, hand it a provider and a symbol, and the first RSI you print is the same number the hosted chart draws.

npm install algobeam-ts
  1. 1Install the package — compiler, runtime and CLI ship together.
  2. 2Point a runtime at a provider, a symbol and a timeframe.
  3. 3Load bars, run an indicator, read the last value.
  4. 4Swap the built-in indicator for your own .algo file.
quickstart.ts
import { AlgoBeamTS, Provider, indicators } from 'algobeam-ts'

// provider, symbol, timeframe, how many bars of history to load
const algobeam = new AlgoBeamTS(Provider.Binance, 'BTCUSDT', '1h', 100)

async function main() {
  const bars = await algobeam.load()
  const last = bars.at(-1)
  if (!last) throw new Error('provider returned no bars')

  const rsi = await algobeam.run(indicators.rsi({ length: 14 }))
  const ema = await algobeam.run(indicators.ema({ length: 50, source: 'close' }))

  console.log('bars loaded  ', bars.length)
  console.log('last close   ', last.close.toFixed(2))
  console.log('rsi(14)      ', rsi.last.toFixed(2))
  console.log('ema(50)      ', ema.last.toFixed(2))
  console.log('regime       ', rsi.last > 55 && ema.rising ? 'expansion' : 'neutral')
}

main().catch((error: unknown) => {
  console.error('[algobeam] run failed', error)
  process.exitCode = 1
})

Results are deterministic: the same bars and the same inputs produce byte-identical output in the CLI, in a browser tab and on the hosted charts.

[ 03 ]Changelog

What shipped, and when

Platform and runtime releases share one version line, so a chart feature and the AlgoBeamScript function behind it always land together.

  1. v4.2.0Latest

    Market Replay goes tick-accurate

    • addedMarket Replay now steps through intraday ticks rather than one-minute aggregates, with a scrubber that snaps to session boundaries.
    • addedReplay sessions record every interaction and export as an annotated review file you can reopen or share with a mentor.
    • improvedThe heatmap hover card now carries the next earnings date, the free float and a 30-bar sparkline.
    • improvedScreener results stream in progressively instead of waiting for the full result set, cutting perceived latency roughly in half.
    • fixedAlerts scheduled on the weekly close no longer fire twice for symbols that trade a half-session on Friday.
  2. v4.1.0

    Portfolio exposure and correlation clustering

    • addedPortfolio tracking gained a correlation cluster view that groups open positions by 60-day return correlation.
    • addedContribution breakdown attributes month-to-date performance by position, sector and strategy tag.
    • improvedRisk Console limits can now be scoped per strategy tag instead of applying only account-wide.
    • fixedFixed a rounding error that displayed sub-penny FX positions as zero-value rows in the exposure table.
  3. v4.0.0

    Quant v4 — chart-aware reasoning

    • addedQuant now reads the indicator stack on your active chart and cites the specific series it used in every answer.
    • addedAnswers link back to the underlying bars, prints and filings, so any claim can be checked in one click.
    • addedFollow-up questions keep the chart context for the whole thread rather than resetting each turn.
    • improvedMedian response time on a 5-year daily series dropped from 4.1 s to 1.6 s.
    • fixedQuant no longer answers questions about symbols outside your entitled data set; it now says so plainly instead.
  4. v3.7.0

    Hosted AlgoBeamTS workers

    • addedUltimate and Enterprise plans can deploy an AlgoBeamScript strategy to a 24/7 hosted worker with one command.
    • addedWorker runs emit structured signal payloads to signed webhooks with replay protection and per-endpoint secrets.
    • improvedThe AlgoBeamTS CLI now reports a diff when a local run and a hosted run disagree on any bar.
    • fixedFixed a daylight-saving edge case that shifted the session clock by an hour for European instruments.
  5. v3.6.0

    Screener query language

    • addedA compact query language replaces nested filter menus: type conditions, save the preset, share the string.
    • addedThirty-one new filters covering short interest, options skew, insider flow and index membership.
    • improvedAny screener preset can now be promoted to an alert or a backtest universe without rebuilding it.
    • improvedPresets sync across seats on team plans and carry an ownership label.
    • fixedSector filters no longer exclude dual-listed names whose primary listing sits outside the US.
  6. v3.5.0

    Walk-forward and Monte Carlo in the backtester

    • addedWalk-forward analysis with configurable in-sample and out-of-sample windows, reported side by side.
    • addedMonte Carlo trade resampling produces a distribution of outcomes instead of a single equity curve.
    • improvedSlippage now scales with the bar range rather than using a flat per-trade assumption.
    • fixedTrades that could not have been filled at the requested size are flagged rather than silently accepted.
  7. v3.4.0

    Strategy Builder

    • addedA block graph editor for entries, filters, exits and position sizing that compiles directly to AlgoBeamScript.
    • addedGenerated source is editable by hand at any point, and edits round-trip back into the graph where possible.
    • improvedIndicator parameter panels now show the historical range of each input alongside its default.
    • fixedUndo history no longer resets when switching between the graph view and the source view.
  8. v3.3.0

    Faster charts and a rebuilt alert pipeline

    • addedAlerts can bundle several conditions into a single rule so one setup produces one notification.
    • addedSMS delivery for Ultimate, with per-rule quiet hours in your local timezone.
    • improvedChart rendering moved to an incremental pipeline: a 10-year daily series now paints in under 120 ms.
    • improvedMedian alert delivery latency fell from 5.8 s to 0.9 s measured from bar close.
    • fixedCrypto alerts no longer miss the first bar after a venue reconnects.

Ready when you are

Read a page, then go build it.

Every example in these docs runs unchanged in the CLI, in the browser and on the hosted platform. Install AlgoBeamTS and check the numbers yourself.

  • No card required
  • Cancel anytime
  • 14-day Ultimate trial