Nyxina — Crypto Watchlist Screener

An honest crypto watchlist — predicted today, graded in public tomorrow. The score you see is the real, measured hit rate, never a made-up '99'.

A ranked list predicting which crypto perps are most likely to touch +10% or more in the next 24h — each with an honest, calibrated probability. And +10% is the floor, not the ceiling: we keep grading every pick at 48h and 72h, where the real winners tend to run much further. It's a forecast, not a report of what already moved — a daily screener for traders, not a live ticker for scalpers.

Pick the edition that fits your timezone — 00:00 / 06:00 / 12:00 / 18:00 UTC (Asia morning → US afternoon). Each is its own independently backtested, separately calibrated list, refreshed once a day at its anchor. It's ready a few minutes later, so pull it any time from :30 past the hour (00:30 / 06:30 / 12:30 / 18:30 UTC) — no need to check before then. Same engine, four snapshots of the day; grab the one nearest when you trade.

Not a scrape-and-rank feed: every score comes from a researched, back-tested model — out-of-sample across every market regime, calibrated so the number is the real hit rate.

🎁 Free while we launch. Pull it as much as you like, no charge — we're building the public track record first. Pricing arrives later (you'll see it here when it does), so early users get in free.

Why Nyxina

• 🚫 Not a top-gainers list. That's free on any exchange — and it's backwards. Our model gives a coin's 24h pump almost zero weight and penalizes ones already stretched from their lows; it ranks on whether +10% is likely next — from volatility and each coin's own history of moving. So a red coin near its low routinely outranks a green one that already ran, and a top gainer only makes the cut if the math still likes it.
• 📋 No code, no setup. Read it as a table right in your browser, or export to Excel / CSV in one click — or Schedule it to run itself daily. Builders get clean JSON/API too. Same data, your format.
• 🎯 Calibrated. A coin showing 38 hit +10% in ~38% of similar past cases. Nothing here shows 99 — other screeners' "99" is a marketing number.
• 🧾 Receipts. The 00:00 edition is graded every day, in public, on the Telegram ledger — misses included, nothing deleted — and each day's full list is sha256-committed when posted, so the next-day reveal provably matches (no quietly editing history). That daily public scorecard is the standard of proof; every edition runs the same engine, each calibrated to its own anchor.
• 📊 Audited. Out-of-sample across every regime, delisted coins included.
• 🧠 Model-driven, kept honest. A machine-learning model — not hand-tuned rules — recalibrated against real outcomes as the live record grows. The backtest never stops.
• 💸 Cheap, on demand. A fresh list lands once a day per edition — pull one when you're about to trade and casual use costs next to nothing. No reason to spam it; one daily glance keeps you on top of the board.

Who it's for

• Busy traders. One glance a day instead of twelve chart hours. Check the list with your morning coffee, pick, size, set your exit clock — then go live your life. The math favors patience, not screen-watching: winners typically need hours to a day, not minute-by-minute babysitting.
• Beginners. Don't know where to start? Use the daily list as your watch-and-learn feed. The graded history shows what a 38 really means, and the risk tags teach you what danger looks like before it costs you.
• Builders. Clean JSON via API — pipe the list straight into your own bot, dashboard, or sheet.

No code needed — and why it's on Apify

You don't need to be a developer:

1. Click Start — a simple form (edition/timezone, list size, tags on/off).
2. Read the results as a table, or export to CSV / Excel in one click.
3. Want it daily at your hour? Hit Schedule and it runs itself.

Builders get the same data as clean JSON / API — pipe it into a bot, sheet, or dashboard. Same data, your format.

Why Apify, not our own website? Because you shouldn't have to trust us. Apify is an established, public automation platform — the screener runs there in a sandbox, with access and billing handled by a neutral third party, so you never hand details to, or run code from, an anonymous site we control. It's a public listing (run stats and reviews in the open), and the track record is fully public: every pick is graded the next day on the free Telegram ledger — misses included — and each day's list is sha256-committed when posted, so the reveal provably can't be edited. Don't trust, verify.

The numbers

Tier	+10% in 24h	in 72h	random coin
Top pick	65%	75%	11%
Top 3	64%	74%	11%
Top 10	52%	66%	11%
Top 30	41%	58%	11%

The edge held in every regime we tested. And +10% is the floor we grade, not a cap — winners don't stop there (median on the sharp end tops ~+30% within 72h, 1 in 10 doubles). The odds are measured from each edition's snapshot price (00:00 / 06:00 / 12:00 / 18:00 UTC — the anchor you're viewing, not always midnight), so getting in below it only widens your gain.

Board read. Each edition carries a one-line board signal — how strong today's board is. We rank the top 10 every day, whatever the regime — so on a quiet board those 10 are inherently weaker and more of them will miss (~41% hit vs ~56% on a strong board). The badge says so up front: size lighter when it's quiet, lean in when it's strong. We'd rather flag a weak day than hide it — the misses on a quiet day are expected, not a surprise. Free on every surface.

We use it too

This isn't a scorer built only to sell — the same engine runs in our own live crypto trading every day; we wouldn't stake our own money on a number we didn't trust. Honesty cuts both ways, though: this is the watchlist, not our exact trades — our live trading adds its own entry and risk filters on top (some high-scorers we watch but don't take), and the edge is in how you enter, exit, and size, not the score alone. That entry read is what the per-coin tip hands you. So you get the genuinely hard part — where to look, with real odds — and we stay upfront that the trade is yours.

Each row

rank · symbol · score (calibrated %) · price_at_score · tags (risk chips: hot, stretched, beaten-down, falling, ...) · tip (a one-line entry read) · venues (also on Bybit / OKX / Hyperliquid / Coinbase) · day (the edition's UTC anchor)

Tags warn, they don't censor — a hot coin with a warning chip can still be the move of the day. Read them as context.

The tip is a soft, plain-language read on entry timing — the gap between "likely to move" and "a good entry right now" (the same thing our own trading weighs). It's guidance, not a graded number: the score is the measured edge, the tip is just context. A clean coin gets a green light; a coin that's already running gets a "don't chase" nudge. Your call, always.

Use it right

1. A watchlist, not a buy list — a daily shortlist to watch, never a "buy these." We give you where to look and the real odds; the entry, exit, and sizing are yours.
2. Spread your picks — 63% still misses 1 in 3.
3. Top rank ≠ best trade — check the tags.
4. Survive the dip — winners often go −5..−15% first. Size for it.
5. Give it a day — median winner needs ~11h; 3 in 10 need more than a day. Selling red at hour 2 donates winners.
6. Don't FOMO — enter cold. The odds are measured at our snapshot price, so enter near it or on the dip and find your moment; chasing a coin that already ran another leg pays up and wrecks the very odds you came for. And red ≠ missed — fresh-and-red is the designed shape, not a failed trade.
7. Mind the macro — then use it. The model reads each coin's own behavior, not the headlines, so it can't see a Fed print or a market-wide BTC flush coming — in a broad risk-off the hit rate dips (the ranking holds, the regime doesn't). So size down or sit out the falling knife. But a crash isn't the end of the list — it's the setup: flushes reset coins to oversold, and the rebound is where the strongest names rip hardest. We predict coins, not crashes — yet timed right (once it's basing, not mid-drop), a flush is fuel, and this is the list that tells you which names to load.

---

Crypto derivatives are extremely risky. Not financial advice. Past performance doesn't guarantee future results. DYOR.