A tier is not just a quota. It decides which pipeline runs — how many of the eight checks execute, whether plain-language translation is available, and how much history you can test against. The limit is enforced in the backend, so a request that slips past the interface is refused or produces a result shaped for its tier.
Free
Built to demonstrate the mechanism cheaply, not to be anyone’s final word.
- Unlimited tests with the standalone Null-info diagnostic. They don’t count as audits.
- Unlimited reports from Read my script, which reads a pasted Pine script and says line by line what it does and whether this platform can validate it. A report runs no backtest, so it never counts as an audit — on any plan. There is a rate limit rather than a quota, and the script isn’t stored. The prompt at Translate your script is public: no account needed to read or copy it.
- 3 audits a month. An audit is one run through the validation stack, and a run counts whether it clears or not. A run that doesn’t run — refused before it starts, say because the data doesn’t cover its dates, or stopped by an error — doesn’t count. The count resets on the 1st of each month (UTC), and the Billing page shows how many you’ve used.
- One symbol per audit. Several symbols in one run are part of Pro; the strategy builder lets you pick one.
- Six of the eight checks: leak audit, twin control, null information, pre-registration (default floors only), temporal halves, cost realism — the ones whose absence would falsify the trust mechanism.
- Every asset class, over the last two years: a run’s date range has to start no earlier than the same date two years ago. The strategy builder only offers ranges inside that window.
- Timeframes of one minute and slower. Sub-minute bars are part of Pro.
- The DSL, the full indicator library, immutable run history, and deflated-Sharpe accounting.
- The raw scorecard, with the two checks it does not run named on it. No plain-language attribution.
- No forward registration. If you registered strategies on Pro, they stay on record and keep accruing after you move to Free, and you can still withdraw them.
Audit pass, coming soon
Not on sale yet. A one-off purchase for a single strategy, meant for one answer before a big decision. It will include:
- Up to three runs of that strategy within 60 days, not counted against Free’s three a month.
- All eight checks, with diagnostic attribution, ablation and trade diagnostics.
- Full history, on one symbol. No sub-minute bars.
- The prop challenge simulator and the trade chart for that strategy.
- A PDF report, and 60 days of forward registration for that strategy.
It won’t include AI drafting or the AI scorecard read-out, and it doesn’t count toward Pro.
Pro
- All eight checks, including placebo and cluster-adjusted significance.
- Audits with fair-use limits, each over several symbols if you like.
- Full history and sub-minute bars.
- Plain-English descriptions compiled to a spec (still with the mandatory confirmation step).
- Diagnostic attribution with case-study links, AI scorecard read-outs, and user-committed floors.
- The interactive trade chart and trade-by-trade comparison.
- The prop challenge simulator.
- Forward registration and its scoring, sub-minute bars included.
Research, later
Research isn’t available yet. At $129/month it will add unlimited custom Python indicators and an API — once the environment that runs your code is production-grade. Pro will get a capped number of custom indicators at that point too.
Billing
Free is $0 with no card. Pro is $39/month or $349/year. Payments go through our payment provider, which handles tax, invoices and cancellation from its customer portal. Your plan is whatever your subscription says; there is no way to set it by hand. Manage it on the Billing page.