Frequently Asked Questions

Million Candles is a chart and market analysis platform for active traders. Upload any chart screenshot and get a structured read on the setup — direction, entry, stop loss, two take-profit targets, R:R and a confidence score — in seconds. It also includes an Options Scanner, Live Tracker for real-time structural analysis, Asset Intelligence deep-dives, a Paper Portfolio for forward testing, live news and more. The analysis is powered by AI trained on price action and Smart Money Concepts, with every level cross-checked against live market data.

No. The free Explorer plan is a good starting point, and the Learner plan includes a full Learning Path explaining charts, risk management and how to read the signals, all in plain English.

Chart analysis works on virtually any instrument — US futures (MNQ, MES, NQ, ES), global indices (NIFTY, FTSE, DAX, Nikkei), forex, crypto, stocks on any exchange, commodities and ETFs. Asset Intelligence and Live Tracker currently focus on US and Indian equities/indices, with more asset classes on the roadmap.

Create a free account (no credit card needed), then go to Scan Mode and upload a screenshot of any chart. The Explorer plan gives you 3 scans a day for 14 days — enough to properly test whether the tool fits how you trade, including multi-timeframe comparisons. After that, scanning requires a paid plan, but your account stays active: Live News, Market Movers, the Daily Call, the Macro Calendar and the calculators remain free indefinitely.

Million Candles is installable as a Progressive Web App (PWA) — add it to your home screen from your mobile browser and it behaves like a native app, including push notifications for alerts.

Any modern browser — Chrome, Safari, Firefox, Edge — on desktop or mobile. No installation required to use the core platform; the PWA install is optional for a more native feel.

Upload a chart screenshot from any platform (TradingView, your broker, anywhere). The AI reads the price structure — swing highs/lows, breaks of structure, fair value gaps, liquidity sweeps — and returns a structured trade idea: direction, entry, stop loss, two take-profit levels, risk:reward ratio and a confidence score, with a plain-English explanation of the structural factors behind the call.

Most AI chart tools are chat-based — you get a conversational response, not a structured, consistent output. Scan Mode uses deterministic, factor-based scoring (real structure detection, not just AI vision guessing) so the same clean setup scores consistently, not differently each time you ask. You get exact numbers — entry, stop, two targets — not a paragraph you have to interpret.

Yes — uploading a second chart of a different timeframe for the same instrument gives the AI genuine multi-timeframe confirmation, which unlocks the full confidence range rather than the single-timeframe adjustment applied to a single upload.

Scan Mode checks whether you’ve recently scanned the same ticker and, if so, compares the two reads. Genuine agreement between the two scans is treated as real multi-timeframe confluence and explained clearly; disagreement is flagged honestly rather than hidden, so you know if your own reads are conflicting before you trade either one.

Scan Mode analyzes any chart image regardless of instrument or exchange — the AI reads the price action itself, not a live data feed tied to a specific market. You can scan US futures, Indian indices, forex pairs, crypto, or any other chart you can screenshot.

Explorer (free): 3 chart scans per day for 14 days, then the scanning trial ends. Learner and above: unlimited scans. The 14-day window is deliberately long enough to test multi-timeframe comparisons, which need more than one scan to demonstrate at all.

It measures how many structural factors (break of structure, fair value gap, liquidity sweep, key level, risk:reward, multi-timeframe confirmation) align in support of the setup, weighted and summed from a fixed formula — not a black box. It is NOT a probability of profit. A high score means the chart structure is clean and factors agree; markets remain uncertain regardless of score.

The confidence score reflects structural clarity, not a guarantee of profit. A high score means the chart structure supports the idea; markets are uncertain and any setup can lose. Use it as one input alongside your own judgement and risk management — never as a substitute for it.

The score is a genuine sum of weighted factors — structure alignment, risk:reward, volume, multi-timeframe confirmation — not inflated by default. A setup missing a genuine second-timeframe confirmation, showing internal conflict between factors, or with a marginal risk:reward will score lower even if it looks clean at a glance. The scoring breakdown under each result shows exactly which factors added or subtracted points.

No. It means the AI’s structural read is internally consistent and well-supported by what it can see in the chart. It has no knowledge of your account size, risk tolerance, other open positions, or news events outside the image. Confidence is about clarity of structure, not a recommendation.

Yes — the Daily Call posts one real, timestamped market call every day, and grades the outcome the same day. Nothing is deleted or hidden, win or lose. It’s the most honest way to see how the underlying analysis actually performs over time, and no login is required to view it.

Yes, and this is stated plainly rather than downplayed. AI output is probabilistic and can misread a chart, miss context, or produce an incorrect level. Always verify levels against your own charts before acting on any analysis.

Generally, yes — confidence is a straight sum of how many structural factors align, so a 90+ means more factors agreed (break of structure, fair value gap, liquidity sweep, multi-timeframe confirmation, clean risk:reward) than a 65 does. But it’s not a probability of winning: a 65 with genuinely clean structure can still work out, and a 90 can still lose. Treat the score as “how much of the structural picture agrees,” not “how likely this is to profit.”

Scan Mode reads the chart fresh every time rather than caching a stale result, so new price action between scans can add or remove structural factors — a fair value gap gets filled, a new swing point forms, a liquidity level gets swept. If you scan the same ticker twice at different timeframes close together, the two reads are compared directly and any disagreement is flagged rather than hidden.

Every call — Daily Call, Scan History outcomes, Paper Portfolio — is timestamped and locked at the moment it’s made, before the outcome is known, then graded afterward against what actually happened. Internally, models are validated the same way: out-of-fold testing (checking a result only against data the model didn’t see while fitting) and minimum-sample gates before any statistic is shown at all, so a handful of lucky trades can’t get dressed up as a real pattern.

Choppy, range-bound price action with no clear break of structure; a single-timeframe scan with no second timeframe to confirm it; conflicting signals (say, a bullish structure break paired with weak risk:reward); or a setup already extended far past any clean invalidation point. The score is reporting less structural agreement, not being cautious for its own sake.

From the chart’s own structure, not a fixed percentage or dollar distance — the stop sits beyond the level that would actually invalidate the setup (a swing point, liquidity level, or structural break), and targets are drawn from the next meaningful structural levels price would need to clear. That’s also why risk:reward varies setup to setup instead of always landing on a round number.

The Daily Call keeps a full public history — every call ever made, graded, nothing deleted, no login required to check it. Your own Scan History does the same for your personal scans, with outcome tracking against real price data. Track records for newer, feature-specific tools are being built the same way — held back from public display until there’s enough graded history for the numbers to actually mean something, rather than publishing thin or cherry-picked stats.

A real-time, continuously-updating structural analysis of any ticker — the same deterministic engine as Scan Mode (structure breaks, fair value gaps, liquidity sweeps, multi-timeframe confluence), but running on a live feed rather than a single uploaded screenshot, refreshing automatically while you watch it.

It depends on the instrument and your setup. For US stocks and crypto with a connected Alpaca account, data is genuinely real-time (free IEX feed). For futures and forex, free real-time exchange data isn’t available anywhere without paying real exchange licensing fees — Live Tracker instead offers an optional free TradingView webhook bridge: you set one alert on your own TradingView chart, and its price updates feed into Live Tracker directly. Without either of those connected, data defaults to Yahoo Finance, which is delayed up to 15-20 minutes — this is disclosed directly in the app, not hidden.

Yes — Live Tracker supports a split view with two independent tracked instruments side by side, each with its own timeframe and refresh cycle.

Yes, if you have a broker connected — the same broker order-staging used elsewhere in the app is available directly from a live-tracked setup, so you can act on a real-time read without switching pages. You always confirm every order yourself.

Live Tracker is currently being refined for full release and access may be limited while real-time data reliability is finalized — check the app for current availability.

It scans across whole sectors or validates a specific trade idea you already have, ranking call/put setups by an AI-computed opportunity score that factors in live options data, IV rank, put/call ratio, max pain, earnings timing, analyst sentiment and breaking news — not just raw numbers.

Market Scanner sweeps an entire sector or watchlist (e.g. Nifty Blue Chips, US Tech Giants) and surfaces only the setups worth looking at. Trade Validator takes a specific ticker, strike and expiry you already have in mind and checks it against live news, policy announcements, analyst ratings and options flow before you commit.

Far-dated options ideas beyond the nearest expiry — LEAPS (long-term options, typically 6+ months out) and Quarterly Plays (roughly one-quarter horizon), for traders positioning for a longer thesis rather than a near-term move. These use real far-dated options chains, not the nearest-expiry data used elsewhere.

Implied volatility at several points along the options chain—near-term, ~1 month, ~2 months, ~3 months out—rather than just the nearest expiry. When near-term IV is notably higher than longer-dated IV (“backwardation”), it usually means the market is pricing in a specific near-term event like earnings.

Seven global markets: USA, India (NSE/BSE), UK, Europe, Australia, Japan, and Global Indices. The AI adapts its analysis to each market’s expiry structure, lot sizes and institutional behavior — for example, US weekly options expire Fridays while Indian NIFTY/BANKNIFTY expire Thursdays.

Yes — beyond whether to enter a setup, the AI suggests how to size the position relative to your overall portfolio, not just a binary entry signal.

A deep-dive dashboard for any ticker: analyst ratings and price targets, fundamentals, insider activity, and — for US tickers — a full options intelligence suite (IV rank, put/call ratio, max pain, open interest by strike, IV term structure). Available on Learner and above.

Shows SEC Form 4 filings — when company officers, directors, or 10%+ owners buy or sell their own stock — for the last 120 days, via SEC EDGAR. It specifically highlights genuine open-market buys/sells, since most Form 4 filings are actually stock grants or tax withholding, not a real bet on the stock.

Real, legally-mandated disclosures of what US Senators have traded, under the STOCK Act, for the last 2 years. This is Senate-only, not the full Congress — there’s currently no equally reliable free source for House disclosures. Amounts shown are the disclosed range, not an exact figure, since Senators only report brackets.

Three additional SEC-sourced signals beyond insider trading: 8-K material events (classified into plain English using SEC’s own official codes), 13D/13G ownership stakes (distinguishing genuine activist intent from passive index-fund positions), and S-3/424B5 filings (a real dilution-risk signal when a company registers to potentially sell more shares).

How a stock has performed against the broad market index (S&P 500 or NIFTY 50) and its sector, across 1, 3, and 6-month windows — shown in percentage points of out- or under-performance, not just raw price change.

Key metrics for a stock — price, market cap, P/E, day change — shown side by side with a few close industry peers, so you can judge whether a valuation is actually cheap or expensive relative to comparable companies, not in isolation.

The next earnings date, the stock’s real EPS beat/miss track record, and — computed from actual historical price data, not estimated — how much the stock has genuinely moved in the days following its last several earnings reports.

Yes — Relative Strength and Peer Comparison both work for Indian tickers (via curated NSE sector mappings), alongside stock fundamentals. Options intelligence (IV rank, max pain, IV term structure) and the SEC-sourced features (Insider Activity, Senate Trades, Corporate Filings) are US-only, since free real-time NSE options data and SEC-equivalent Indian filing data aren’t available the same way.

It lets you forward-test setups as hypothetical trades — no real money. Each is automatically graded against real price history to show whether it would have hit target or stop first, so you build an honest track record before risking capital. Available on Learner and above.

A full log of every chart you’ve scanned, with outcome tracking (hit target / hit stop / pending) so you can see how your own past setups actually played out, and calibrate your trust in the analysis over time rather than just trusting the confidence score blindly.

Yes — open, journaled positions are periodically re-checked, and you get an invalidation alert if the underlying structure that supported the original call breaks, so you’re not holding a position the AI would no longer support.

One real, timestamped market call posted publicly every day — entry, stop, target, full reasoning — graded the same day, with nothing ever deleted or hidden. It’s a public, ongoing record of how the analysis actually performs, viewable with no login required.

The Daily Call currently focuses on US markets. Under SEBI regulations, Million Candles does not publish daily market calls for Indian instruments to Indian users, in line with India-specific compliance requirements.

It’s designed to be checkable, not just claimed — every call and its graded outcome stays visible permanently, win or lose, so you can verify the actual hit rate yourself rather than take a marketing claim at face value.

You connect a US broker and set your rules in Risk Shield (confidence threshold, max contracts, daily loss limit, trading hours). When a TradingView alert fires, the system re-analyzes the setup fresh — not the stale snapshot from when your strategy triggered — and only executes if the fresh confidence score clears your threshold and every risk rule passes.

Typical webhook bots fire an alert straight to the broker even if the market has moved in the seconds since the signal fired. AutoPilot puts an AI checkpoint between your strategy and your money — re-confirming live structure, momentum and key levels before any order goes out, and logging the decision either way so you always know why a trade was placed or skipped.

No — in accordance with SEBI’s algorithmic trading framework, automated execution is not offered for Indian brokerage accounts. Indian users get the full AI analysis suite but place every order manually through their own broker’s window (e.g. Zerodha Kite), with sessions requiring fresh authentication daily.

US futures and equities via Rithmic and Alpaca, with TradeStation and others in active discussion. Supported brokers change over time — check Launch Pad for the current list before subscribing, since broker availability isn’t grounds for a refund per the Terms of Service.

Yes — Risk Shield rules are checked before every single execution, built with common prop-firm constraints (like Apex or Topstep-style daily loss and drawdown limits) in mind. It’s a convenience tool operating on a best-efforts basis, not a compliance guarantee — you remain solely responsible for knowing and following your specific firm’s rules.

The signal is logged with the reason it didn’t pass, and no trade is placed. Every decision — executed or skipped — is recorded in a full audit trail so you always know exactly what happened and why.

India: Zerodha (Kite Connect, live) and DhanHQ (OAuth-connected, live). US: Rithmic (futures, live), Alpaca (stocks, live), with TradeStation and NinjaTrader in active discussion.

Broker API credentials and OAuth tokens are encrypted with AES-256-GCM before storage and decrypted only at the moment of order transmission — never stored in plain text, never shared with any third party. Connections are trade-only: Million Candles cannot withdraw funds, and for manual execution you confirm every order yourself.

For manual execution (all Indian brokers, and US brokers outside AutoPilot), no — you review and confirm every single order. AutoPilot (US-only) executes automatically, but only within the exact rules and limits you configured yourself in Risk Shield beforehand.

Submit a request on the public Roadmap page — broker integrations are prioritized based on demand, and every submission is reviewed by the team.

Via Kite Publisher, which runs entirely in your browser on Zerodha’s own official pages — your Zerodha credentials never touch Million Candles’ servers, and every order is confirmed on Zerodha’s own order window.

Where you configure the rules AutoPilot enforces: confidence threshold, max contracts per trade, daily loss limit, trading hours, and prop-firm-style drawdown limits — checked before every single automated execution.

No — it’s a best-efforts convenience tool, not a compliance guarantee. You remain solely responsible for knowing and complying with your specific firm’s rules; this is stated directly in the Terms of Service and Risk Disclaimer, not buried in fine print.

The Explorer plan gives you 3 chart scans per day for 14 days, no credit card required. After the trial, chart scanning needs a paid plan — but Live News, Market Movers, the Daily Call, the Macro Calendar and the calculators stay free with no time limit.

Unlimited chart scans, Asset Intelligence, Paper Portfolio, full Scan History with outcome tracking, and the complete Learning Path.

Everything in Learner, plus the Options Scanner, broker execution (stage orders to your own connected account), and push/invalidation alerts.

Everything in Active Trader, plus rules-based automated execution, the TradingView webhook bridge, and prop-firm protection. US accounts only.

Yes — upgrades take effect immediately with Stripe automatically pro-rating the difference; downgrades take effect at your next renewal.

Yes. Cancel any time in Settings → My Plan with no penalties — access continues until the end of your current billing period, then it simply won’t renew.

Your first subscription payment is eligible for a full refund within 48 hours, provided your account has run fewer than 5 chart analyses and no AutoPilot executions have occurred, once per customer ever. See the full Refund Policy for exact terms.

Pricing is shown in your local currency based on your location, but plan tiers and features are the same globally, with the exception of AutoPilot and the Daily Call, which are US-only per SEBI compliance requirements for Indian users.

For $49/mo on top of any paid plan, you can add 4 more people — family, friends or your trading group — each with their own login, scan history and journal, all under one bill. That works out to about $12 a head.

No — each team member connects and trades their own broker account individually. Accounts are never mixed or shared, only the billing is combined.

Yes, team seats can be managed from My Team at any time.

Million Candles is not registered with SEBI as an investment adviser or research analyst — the platform is educational software, and this is stated directly rather than implied. In line with SEBI’s algorithmic trading framework, AutoPilot (automated execution) is not offered for Indian brokerage accounts, and the Daily Call’s public market calls are not shown to Indian users.

Every order is individually initiated and confirmed by you — through Zerodha’s own Kite Publisher order window, for example — never executed automatically. Broker sessions require fresh authentication daily, in line with SEBI requirements.

Only after Million Candles completes exchange empanelment as an algo provider under SEBI’s framework. There’s no committed timeline for this.

No. Million Candles does not sell personal data and does not use it for third-party advertising. Data is shared only with the processors needed to run the service (Stripe for payments, Supabase for hosting, Anthropic for AI processing, your broker if you connect one), each bound by their own security obligations.

Chart images you upload are processed to generate your analysis and are part of your account’s scan history. You own the charts you upload; Million Candles has a license to process them solely to provide the analysis.

API keys and OAuth tokens are encrypted with AES-256-GCM before storage, decrypted only at the moment of order transmission, and never stored in plain text.

Yes, directly from Settings. Account data is deleted within 30 days of account deletion; encrypted broker credentials are deleted immediately upon disconnection or account deletion.

From Settings, once logged in.

After a grace period, your account is downgraded to the free Explorer plan rather than losing access entirely.

No — accounts are one per person and may not be shared, sold, or transferred, per the Terms of Service.

No. Million Candles is an educational tool only. Nothing on the platform is financial advice, and the platform is not a SEBI-registered research analyst or a US investment adviser. All trading decisions and their outcomes are entirely your own.

Every user gets a unique referral link. When someone subscribes using your link, you get a free month added to your account, with no limit on how many people you refer.

Under Referrals in your account, once logged in.