Skip to main content
PASSFAILINSUFFICIENT-DATA

Test trading strategies. Get an honest verdict.

ChartStrike is an AI-powered strategy platform. Compose a strategy in Research Studio, let a bull/bear/judge debate engine analyze it, and get a statistically-defensible verdict - not the answer you were hoping for.

Honest by construction: deflated-Sharpe, overfitting (PBO), and power checks, with INSUFFICIENT-DATA by default - so a PASS has to earn it.

Start free - no cardSee how the verdict works

Free monthly testing credits · 158 tools · paper trading · API & MCP access

Every config registered to a trial ledgerVariant-hunting taxed, never freeLook-ahead & stale edges blocked by designEvery stat shown with its denominator

The difference

Most backtesters tell you what you want to hear.

Run a strategy enough ways and something will look great by luck. ChartStrike is built to catch that - so a PASS actually means something.

  • Denominators on everything. "64% win rate" means nothing without the count behind it. We always show it.
  • Trial tax. Every variant you try is registered and charged against your evidence - no free re-rolls until something works.
  • Deflated Sharpe & power checks. We discount for how hard you searched, and refuse to grade an under-powered test.
  • Recency & look-ahead guards. A stale or peeking edge returns INSUFFICIENT-DATA, never a false PASS.
Start testing free
Abstract statistical distribution showing a confidence interval band and a pass marker near a threshold, with an insufficient-data zone.
verdict = survives OOS · across regimes · on recent data
Abstract node graph illustration of a visual pipeline builder with composable tool blocks linked by glowing wires.
SENSE → INTERPRET → DECIDE → SIZE → EXECUTE → LEARN

Research Studio

Compose a strategy from 158 tools.

Drag tools into a visual pipeline - six stages, free-form composition. Mix a day-trade trigger with a macro-regime classifier and an options-greeks read in the same strategy. No code required.

1
Sense
price · macro · flow
2
Interpret
regime · features
3
Decide
triggers · rules · AI
4
Size
risk · sizing
5
Execute
entries · exits
6
Learn
verdict · journal
Equities - liveOptions - liveCrypto - pending releaseFutures - pending releaseFX - pending releasePrediction markets - pending release

The debate engine

A bull, a bear, and a judge argue over your strategy.

Three AI advisors stress-test every pipeline from opposing sides. The judge weighs both cases against the evidence and the honesty guardrails - then issues the verdict.

Bull

The strongest case for it working

Surfaces where the edge concentrates, which regimes it wins in, and what makes it defensible.

Bear

The strongest case against

Hunts overfitting, fragile regimes, thin samples, and look-ahead - the reasons it might be luck.

Judge

The verdict, with receipts

Resolves the debate against the statistics and your validation rigor. PASS / FAIL / INSUFFICIENT-DATA - with the denominators that earned it.

Three glowing orbs in a triangle representing bull, bear, and judge roles connected by luminous threads.
house pipelines and API agents face the same judge - nobody gets a free pass

Everything in the box

A full strategy workbench.

Research Studio

The 158-tool pipeline builder with on / off / locked / newly-added tool states and live coverage checks.

Journal

Full run history. Every backtest, every verdict, every variant - searchable and reproducible.

Presets

Start from Strict, Balanced, or Aggressive validation profiles, or save your own. Labels never affect the hash.

Live signals

Promote a PASS to live monitoring. Get signals as your strategy fires, with a kill switch that's always one click away.

CS Assistant

A CS Assistant that helps you compose and refine strategies, grounded in your selected tools and pipeline. Proposals never auto-apply, and every transcript is recorded.

API & MCPdevelopers

Point any AI agent at ChartStrike over REST or MCP. Agents face the same honesty guardrails and trial tax you do.

Paper trading

Run any strategy forward on live data with simulated capital before you risk a cent.

Pricing

Free to preview. Pay for what you run.

Two credit buckets, plus subscription tiers with built-in safety rails. Preview is always free; committed runs spend CS Assistant credits.

Monthly testing credits

free

A fresh allotment of CS Assistant credits every month, included with every plan. These reset monthly, are use-it-or-lose-it, and are spent first.

Preview is always free. Monthly testing credits never roll over.

Balance credits

rolls over

Persistent purchased balance for bigger runs and agents. Balance credits roll over, and larger packs include a volume bonus.

500 ◇ = $5
base pack
2,200 ◇ = $20
+10% bonus
6,000 ◇ = $50
+20% bonus

Free

$0/mo

Get a real verdict, free.

  • 10 CS Assistant credits ◇/mo
  • 1 pipeline slot
  • Core tools + full teaching content
  • Journal & paper trading
  • Live signals & API - paid
Start free
Most Popular

Pro

$39/mo

For serious strategy work.

  • 60 ◇/mo included
  • 5 pipeline slots
  • Full 158-tool catalog
  • Live signals + CS Assistant
  • API & MCP access
Start Pro

Elite

$99/mo

Scale, automation, agents.

  • 200 ◇/mo + priority GPU
  • 10+ pipeline slots
  • Advanced sizing tools
  • Priority compute
  • Higher API / agent budgets
Start Elite

Developers & agents

Point any AI agent at ChartStrike.

A REST API and a native MCP server expose Research Studio, the tool catalog, and the verdict engine. Build agents that compose, test, and ship strategies - under the same honesty guardrails and budget caps as the UI.

Get an API key
POST /v1/pipelines/run
{ pipeline: "momentum-v3",
  rigor: "strict" }
→ verdict: PASS
  n=1736 · conf 97/95 · trial 15
also available over MCP

Stop guessing. Get a verdict.

Build your first strategy in Research Studio and get an honest PASS, FAIL, or INSUFFICIENT-DATA - on us.

Start free - no cardSee pricing