Crypto Fear & Greed + BTC Price, Funding & Risk Signals

Collect the daily Crypto Fear & Greed index (0–100) published by alternative.me, with the provider's classification and a derived regime label for each day. Each run returns up to a year of daily readings plus a summary with the latest value, the change since yesterday and the 7-day average. Use it for sentiment dashboards, market-timing research and extreme-sentiment alerts on Telegram.

Market enrichment is ON by default (enrichMarket: false restores index-only collection). Coinbase Exchange public BTC-USD daily candles add completed UTC daily closes and exact calendar 1/7/30-day percentage returns. Hyperliquid public info adds the mean hourly perpetual funding rate (decimal, not annualized), sample count and completeness, plus current open interest in BTC. No keys or paid sources are needed. Source failures preserve every sentiment row with null market values and enrichmentStatus notes.

Each row adds btc (closeUsd, returnsPct, fundingRateDailyMean, fundingSamples, fundingDayComplete, openInterestBtc, openInterestAsOf, divergence), sentimentPercentile, regimeChanged, and enrichmentStatus. Percentile is the percentage of returned sentiment readings less than or equal to this reading (ties included); it uses only returned readings up to and including that day. sentimentPercentileBasis gives the number of readings used; the basis depends on the requested window. Regime change compares adjacent calendar days; the first day or gaps yield null. Divergence flags fear (<=46) with a positive 7-day return or greed (>=55) with a negative return; otherwise aligned, or null without prices.

Limits: reading D uses the candle stamped D-1, which closes at D 00:00 UTC, and 1/7/30-day returns ending at that close. Funding covers the 24 hours before D 00:00 UTC. Today’s row therefore has price and returns when the sources supply those completed candles. Funding means use available hourly samples, with 24 required for a complete day. Venue retention or missing candles leaves gaps; no forward filling. Open interest is only a current snapshot on today's row, with collection timestamp, never historical backfill. Market data sources may be slow or unavailable. If index collection cannot finish, the run fails clearly; unavailable market enrichment retains index rows with warnings. Signals are descriptive research data, not investment advice.

Billing remains one daily row = one default dataset-item charge, including partial rows; there are no extra market rows or custom charge events.

Quick Start

{"limit": 30, "enrichMarket": true}

Input

Name	Type	Default	Meaning
enrichMarket	boolean	true	Add BTC price, funding and risk signals; unavailable sources yield nulls.
limit	integer	10	Days of history to return, 1–365.
alertExtreme	boolean	false	When true and a Telegram chat is set, sends one alert if the latest day is extreme fear or extreme greed.
telegramBotToken	string	empty	Secret Telegram bot token for alerts; the TELEGRAM_BOT_TOKEN secret environment variable also works.
telegramChatId	string	empty	Chat that receives the alerts.

What it returns

One dataset row per day, oldest to newest:

Field	Meaning
recordKey	Stable key, fng:{date}.
source, kind	Record source and type.
date, timestamp	UTC day of the reading and its Unix timestamp.
value	Index value, 0–100.
classification	The provider's own label, unchanged.
regime	Derived label from extreme-fear to extreme-greed.
sourceUrl	Where the reading was published.
collectedAt	UTC time the reading was collected.

The run summary (OUTPUT) reports the latest reading, the change since yesterday, the 7-day average with its regime, the alert result and any warnings.

Output example

Illustrative record:

{"recordKey":"fng:2026-09-08","source":"alternative.me","kind":"fear_greed","date":"2026-09-08","timestamp":1788825600,"value":22,"classification":"Extreme Fear","regime":"extreme-fear","sourceUrl":"https://api.alternative.me/fng/","collectedAt":"2026-09-08T12:00:03Z"}

FAQ

How fresh is the data? One reading per UTC day, published around 00:00 UTC. Each run returns the newest days available.

How much history is available? The series starts on 2018-02-01; each run returns up to 365 days.

How are the labels chosen? classification is the provider's label, unchanged; regime uses boundaries calibrated on the full published history. A mismatch between the two is reported as a warning, never a failure.

What happens on bad rows? Malformed days are skipped and named in the run warnings; the rest of the run completes.

Price

See the Actor's Pricing tab on Apify for current rates.

For support, contact Apify support and include the Actor run ID.