Backtest your Pine Script
and find out which parameters
actually mattered.

A hosted Pine Script v6 backtester and strategy research workspace for crypto. It fetches Binance klines directly, executes your strategy against them, reports equity, drawdown, Sharpe, Sortino and 40 other metrics, then searches the strategy's own input() parameters with seven algorithms and analyses the recorded trials. The part most backtesters skip is that it can optimise two or three metrics at once and hand back a Pareto front instead of a single winner, and it ranks which parameters actually moved the result.

Seven: grid, brute force, random search, Bayesian TPE, simulated annealing, a genetic algorithm, and multi-objective NSGA-II. An Auto mode picks one of the six single-objective methods for you from the shape of the search space and your budget, and explains its choice. Auto never selects simulated annealing, the genetic algorithm or NSGA-II; NSGA-II is manual-select because the Auto picker has no notion of an objective vector.

Most optimizers collapse every trial into one score and return the winner, which quietly assumes one metric decides. NSGA-II optimises two or three metrics at once and returns the set of parameter sets where you cannot improve one objective without giving up another. You choose two or three objectives from Net PnL, PnL %, Max DD %, Max DD, Profit Factor, Sharpe, Win Rate and Trade Count, and the workspace plots the front, marks the knee point as the best compromise, and lets you load any point back into the strategy. It is a map of trade-offs, not a profitability claim, and it is a snapshot of what was evaluated rather than a converged front.

Not to compute. The chart, optimizer and analysis workspaces are fully available to a guest with no session at all, including complete optimizer runs. An account adds sign-in, a profile and a security tab, and it is where your credit balance and billing details will live once usage metering is switched on. It does not unlock the optimizer, and it does not sync your saved runs, bar cache or optimizer results between machines, because those stay in your browser.

Nothing is priced yet, and this page will say so until it is. IndicaView is a hosted service whose usage will be metered in credits, but the credit plan, its price and how a balance is topped up have not been announced. You can use the workspace today without a paid plan; when credit-based billing goes live it will be described here and in your account, and the metering will be visible before it affects a run.

Candles are fetched by your browser from the public Binance REST API at api.binance.com. Cached bars, backtest runs and optimization results stay in browser storage. Signed-in accounts use Supabase for authentication and store a profile, saved scripts, custom symbols and timeframes, preferences and in-app notifications. Those records are scoped to their owner. There is no telemetry or analytics, and only the publishable key reaches the browser.

No. IndicaView is a research simulator. It never connects to a broker and never places an order. It helps you test a strategy, compare parameter sets and understand which parameters carry the result. It makes no promise of profit.

Not currently, and the interface says so. The results panel carries an explicit "Includes fees? No" disclosure rather than quietly presenting gross figures as net. Separately, indicator-mode Pine scripts draw markers without fills, so only strategy() mode produces real profit and loss numbers, and the interface tells you which mode your script is in.

There is a separate Python command-line tool that runs a trade log through a Monte Carlo simulation to estimate challenge pass probability, expected value and stagnation. It is not part of the browser workspace and does not read your optimization runs directly. It works from a trade log you export, and it resolves its firm directory at runtime rather than hardcoding any particular firm.

Every run records the algorithm, the seed and a job id, and every trial is retained and exportable as CSV or JSON. A reported number is something you can re-derive and inspect rather than something you have to believe. The parameter importance view also grades its own confidence and labels a thin sample as a hint rather than a fact.

It runs Pine Script v6 syntax, but it is not affiliated with, or dependent on, TradingView. The charting and execution layers are Vela and PineTS from LuxAlgo's open-source stack, and strategies are ordinary .pine text files you own and can read in any editor.