How we build every number on this site.

Primary data sources

The overwhelming majority of market data on Cryptopricing originates from CoinGecko’s public API. We chose CoinGecko as our canonical source for three reasons: its methodology is transparent and publicly documented, it aggregates across a wide spread of exchanges rather than relying on any single venue, and its free public API is rate-limited at a level that lets us serve a small independent site without negotiating a commercial relationship. The tradeoff is that when CoinGecko is slow, missing a coin, or publishing a figure that looks unusual, our numbers inherit that - we try to be explicit when the inherited signal is our bottleneck.

Specifically we pull five endpoints from CoinGecko on a rolling one-minute cadence: the coins market list (top 300 by market capitalisation), the per-coin detail endpoint (for description, links, images, supply details), the market chart endpoint (for sparklines and the 1h/24h/7d/30d/1y windows), the categories list, and the global endpoint (total market cap, 24-hour volume, dominance ratios). We paginate through the top 300 coins in three batches so that rate-limit headroom stays comfortable even during traffic spikes.

News comes from three publishers we selected for their editorial coverage of the crypto space: Cointelegraph, Decrypt, and The Block. Each feeds in via public RSS. We refresh the aggregate every ten minutes - news feeds update more often than that, but the marginal gain of polling every two minutes isn’t worth the extra load on the publishers. Admin-curated news posts are merged into the same stream (see Editorial content below).

A small number of coins (AurumX, Selenite, Q-Bit) are notin the CoinGecko index because they trade on smaller exchanges. For these we read the order book directly from Dex-Trade or BiFinance’s public market APIs. These coins are clearly labelled on their pages as originating from those exchanges rather than the aggregate.

Refresh cadence and caching

The site runs an ingest loop that hits our upstream APIs, writes results into a local SQLite database, and serves every public page from that database. We do not proxy calls through to upstream APIs in the request path - the user’s browser talks only to us, and we talk to CoinGecko on our own schedule. This insulates users from rate-limit rejections, keeps response times predictable, and lets us serve last-known-good values gracefully when an upstream is temporarily down.

Different surfaces have different refresh profiles:

• Prices, 24h change, market cap, volume: refreshed every one to two minutes. The "prices updated Ns ago" chip on the home page and the coin detail page tells you exactly how fresh the displayed numbers are, on a rolling basis.
• Historical charts:fetched on demand per timeframe request and cached for five minutes. Switching between 7d and 30d will occasionally trigger a fresh upstream call; that’s expected.
• Per-coin detail (description, links, categories): refreshed every hour. These fields change infrequently so it would be wasteful to refresh them on the same cadence as the price.
• Categories + exchanges tables: refreshed once every few hours.
• News RSS: aggregated every ten minutes.
• Social preview images (OG images on the guide cards): fetched once, cached for a week on success or ten minutes on failure so a broken destination page retries soon without hammering the upstream server.

Every outbound API call is keyed and stored in a generic cache table so that transient failures fall back to the last-known-good value rather than throwing a blank page. When you see stale data, it’s almost always because we’re serving cache from before an upstream hiccup rather than returning an error.

Rank and sort methodology

The "#1, #2, #3..." number next to each coin is CoinGecko’s market-cap rank - we do not recompute rank ourselves. Rank is based on circulating market capitalisation (current price x circulating supply), notfully-diluted valuation or total supply. This means a coin with a small circulating float and a huge total supply ranks lower than its FDV would suggest; we think this is the correct choice for a general-purpose market board, but you can cross-check on the coin’s detail page where we show all three supply figures side by side.

The "Top gainers" column on the home page and the gainers/losers page both use 24-hour price change as the sort key, restricted to coins with a sensible minimum daily volume to exclude illiquid names with volatile quotes. The "Memes" column is filtered to the CoinGecko meme-token category and sorted by 24-hour change; the "Stablecoins" column is filtered to stablecoins and sorted by market cap. The ICO Drops column blends our admin-curated entries with a small curated seed and is ordered by the explicit position field editors set when they feature a project.

We do not support per-user sort preferences in the stored state - sort is purely URL-driven, so a sorted view is shareable by URL and doesn’t leak into the next person’s session.

Computed fields vs raw fields

Most fields on Cryptopricing are verbatim copies of what the upstream API returns. A few are computed by us; we call those out here so you can trust each figure individually:

• 7-day percentage change for exotic coins:computed from the first and last points of the 7d sparkline, because the exchange API doesn’t return a ready-made 7d change field.
• Sparkline tone (green/red/flat): derived from the 24-hour change, not from the sparkline points themselves - so a sparkline can visually rise while still being tinted red if the window ends below where 24h-change measurement started.
• Sidebar "24h high / low":for exotic coins this comes from the exchange ticker’s rolling 24h window; for CoinGecko-tracked coins it comes from the max/min of the 1-day sparkline. These aren’t quite equivalent - CoinGecko’s 1-day sparkline is built from hourly closes, so intra-hour wicks are smoothed out.
• Price predictions: the right-rail forecast card extrapolates the recent percentage change forward and widens a confidence band using 24-hour volatility. When an editor has set an explicit override for a horizon, we display the override and flag the card with an "analyst" pill. Neither the algorithmic nor the analyst forecast is a recommendation to buy or sell.
• Circulating-supply display: shown as a raw count with the coin symbol appended (e.g. "20.02M BTC") rather than as a percentage of max supply, because max supply is often inflated or unknown for coins with emission schedules that extend past 2140.

Exotic coins

Three coins on Cryptopricing live outside the CoinGecko universe: AurumX (ARX) and Selenite (SLNT), both quoted against USDT on Dex-Trade; and Q-Bit (QBT), quoted against USDT on BiFinance. These coins are included because they are tokens held or used by the company (Infraviewer LTD) and we wanted to track them alongside the broader market on the same surface.

Their live data (price, 24-hour change, volume, 24-hour high/low) comes from the relevant exchange’s public market API. We pull a ticker snapshot every minute and a candle history when the chart timeframe requires it. Market cap for these coins is computed as price x display supply - a plausible circulating figure encoded in our exotic-coin definition file, not a verified on-chain count. We mark this with a "price x display supply" hint on the coin page so the number is never mistaken for an audited circulating-supply figure.

Exchange information on these coin pages links directly to the order-book page on the source exchange, not to the project’s marketing site, so clicking "Dex-Trade" takes you to the actual trading page. (That wasn’t always the case - an earlier build pointed both links at the project website; the fix is in place now.)

Editorial content

Two kinds of editorial content appear on Cryptopricing alongside the market data: ICO Drops projects and news posts. Both are managed by the editors through an internal panel and both are clearly attributed.

ICO Drops entries are a mix of projects manually entered by editors (with a logo, status, round, raised/valuation figures, summary, links) and projects scraped from the public icodrops.com listing. When both sources reference the same slug, the editor-managed record takes precedence so edits show up immediately. Featured entries float into the home-page hero column; non-featured ones appear only on /ico-drops. Editors can add, edit, or delete ICO rows at any time and changes propagate on the next page render.

News posts are custom articles written by editors and published with a headline, description, optional body, source label, hero image, and a publish timestamp. They are merged into the public news stream alongside RSS headlines - interleaved by timestamp - so that editorial context can appear next to aggregated headlines without displacing them. Pinned posts float to the top of every news surface. Scheduled posts (future publish timestamps) are invisible until their scheduled time. Editors can attach custom hero images to replace the RSS-supplied ones, and there is a live preview while writing so the card shape on the live site is visible before publishing.

Editorial posts are never sponsored. If we ever accept sponsored content, it will be marked "sponsored" in the source-label slot, not disguised as editorial.

Converter and predictions

The converter page and the sidebar mini-converter both quote a coin against any other coin in our top 50 list plus the three exotics, and against the eleven major fiat currencies CoinGecko publishes conversion rates for (USD, EUR, GBP, CAD, AUD, JPY, CNY, INR, BRL, KRW, CHF). Conversions are computed from the USD-quoted price of each asset rather than from a direct exchange pair - so a BTCETH quote on Cryptopricing is really btc_usd / eth_usd, not the BTC/ETH order-book mid. This matches the behaviour of every aggregator-type site you would compare against, but it means quotes can drift slightly from a market-maker quote on an actual exchange pair when the two USD quotes disagree.

Price predictions shown in the right rail of the coin page are explicitly labelled as an educational estimate and not as financial advice. Algorithmic predictions project the current price forward using the most recent percentage-change reading for each horizon and widen a low/high confidence band proportional to 24-hour volatility. That approach has no meaningful predictive value at any horizon longer than a few days; it exists so that the "what if" conversation is grounded in a concrete number rather than a handwave. Admin-set forecasts override the algorithmic ones and carry an "analyst" pill; these express the editor’s view and, again, are not investment advice.

Known limitations

No data provider is perfect and we’d rather name our limitations than hope nobody notices them:

• CoinGecko coverage:long-tail or newly-listed coins sometimes appear in CoinGecko’s listing a day or two later than they appear on exchanges. Our top-300 window means the 301st-by-market-cap coin is never on Cryptopricing, even if you search for it specifically.
• Exotic-coin market cap: uses an editorial display-supply figure. The number is a plausible estimate, not a verified circulating supply. Treat the "MCAP" column on those coins as illustrative.
• Fiat conversions:pass through CoinGecko’s hourly FX rates. Very fast-moving FX periods can leave the displayed fiat values stale for up to an hour.
• Sparklines: hourly-closing-price series, not tick-by-tick. An intraday wick you remember seeing on a trading view may not show up here.
• News thumbnails: we proxy RSS-provided images through our origin. If a publisher changes their image-hosting setup or removes an image after publication, our cached URL may 404 - we show a fallback tile when that happens, but the headline text is unaffected.
• On-chain fields: we currently do not compute on-chain holder counts, wallet distribution, gas fees, or staking yields. These are on the roadmap but are not in the current data set.

Corrections and takedowns

If something on Cryptopricing is wrong - a stale price that’s clearly not updating, a wrong symbol, a broken exchange link, an article that shouldn’t be there - please email support@cryptopricing.net with the URL and a short description. Corrections that affect a displayed number will typically propagate within one ingest cycle (about a minute) once we push the fix.

For DMCA takedown notices or trademark concerns, please follow the process laid out in our DMCA policy. For everything else legal - GDPR subject-access requests, regulator queries, law-enforcement orders - please use legal@cryptopricing.net so it lands in the right inbox.