Is There a Free Glassnode API Alternative for MVRV?
If you have searched for a glassnode api alternative free mvrv, you already know the shape of the problem. Glassnode's Studio is excellent, but its API is not part of the cheap plans. As of October 2026 its pricing page lists Advanced at $49/mo billed annually with no API access, daily resolution and a personal-use licence, and Professional from $584/mo billed annually with the API as an optional add-on. That is the price of programmatic MVRV from Glassnode.
For a bot that reads a handful of Bitcoin cycle indicators once a day, that is the wrong tool at the wrong price. The indicators themselves (MVRV Z-Score, NUPL, Puell Multiple, Pi Cycle Top, the 200-week MA and friends) are daily, BTC-only series computable from public chain data. We compute all ten from Coin Metrics' daily history and serve them at GET /api/v1/market-intelligence/btc/cycle-indicators on the free tier.
This post answers the three questions a developer asks before switching: which indicators, on which tier, from what date. It does not pretend to replace Glassnode's hundreds of entity-adjusted metrics. If you need SOPR by cohort or hourly realized price, that is still Glassnode. If you need the cycle dashboard most strategies actually gate on, read on.
Which Indicators Does the Free Endpoint Serve?
Ten, each a daily series oldest-first, each entry carrying price (BTC/USD), date, time (unix seconds) and timestamp (unix ms) plus the indicator's own fields. The names match what CoinGlass served until 2026-07-22, so archived coinglass_cycle_indicators snapshots join on them.
| Indicator key | Fields | What it says |
|---|---|---|
mvrv_zscore | mvrv, mvrv_zscore | (market cap - realized cap) / std of market cap up to that day. Expanding window, no lookahead. |
nupl | nupl | Net unrealized profit/loss = 1 - 1/MVRV. |
puell_multiple | puell_multiple | Daily issuance USD / its 365-day mean. >4 has marked tops, <0.5 bottoms. |
pi_cycle_top | ma_111, ma_350_mu_2, ma_110 | A top prints when the 111-day MA crosses above 2x the 350-day MA. |
golden_ratio | ma_350, ma_350_x3 ... ma_350_x21, accumulation_high_1_6, low_bull_high_2 | Fibonacci multiples of the 350-day MA. |
two_year_ma | moving_average_730, moving_average_730_multiplier_5 | 2-year MA and its 5x band. |
two_hundred_week_ma | ma_200w (also moving_average_1440) | The 1,400-day mean that has held every bear-market low. |
ahr999 | ahr999_value, average_price, growth_valuation | <0.45 accumulate, 0.45-1.2 DCA, >1.2 late cycle. |
bitcoin_profitable_days | profitable_percentage, side | Share of past days that closed below this one. |
stock_to_flow | stock_to_flow, next_halving | Supply / trailing-365-day issuance, plus days to the projected halving. |
Three CoinGlass-only series are not served and never will be, because no free source exists: rainbow_chart, bitcoin_bubble and bull_market_peak. Asking for one by name returns 410 Gone, pointing at the archive where their CoinGlass history up to 2026-07-22 still lives.
The Call, and What Comes Back
One request returns all ten. The default is the last 30 days per indicator; days=0 returns everything back to 2010-2014 (about 8 MB, so send Accept-Encoding: gzip).
curl "https://cryptodataapi.com/api/v1/market-intelligence/btc/cycle-indicators?days=30" \
-H "X-API-Key: cdk_live_your_key" -H "Accept-Encoding: gzip" --compressedAn example excerpt (values illustrative, shape exact):
{
"available": ["puell_multiple", "pi_cycle_top", "golden_ratio", "two_year_ma",
"two_hundred_week_ma", "ahr999", "bitcoin_profitable_days",
"stock_to_flow", "mvrv_zscore", "nupl"],
"indicators": {
"mvrv_zscore": [
{"date": "2026-10-22", "time": 1792627200, "timestamp": 1792627200000,
"price": 71250.0, "mvrv": 1.84, "mvrv_zscore": 1.21}
],
"nupl": [
{"date": "2026-10-22", "time": 1792627200, "timestamp": 1792627200000,
"price": 71250.0, "nupl": 0.456}
]
}
}Need one series only? GET /api/v1/market-intelligence/btc/cycle-indicators/{indicator} takes the same days parameter:
curl "https://cryptodataapi.com/api/v1/market-intelligence/btc/cycle-indicators/puell_multiple?days=0" \
-H "X-API-Key: cdk_live_your_key" --compressedTwo timing facts matter for a daily job. The series refreshes every 6 hours and the newest day is usually yesterday, because Coin Metrics publishes a day after it closes. Every value is as-of its own day: nothing is revised backwards, so a backtest over the days=0 pull is point-in-time honest.
MVRV Zones Without the Arithmetic: /on-chain/dormancy/btc
If what you want is a cycle label rather than a raw ratio, GET /api/v1/on-chain/dormancy/btc returns the current MVRV already classified into five zones: capitulation below 1.0, accumulation 1.0-1.5, neutral 1.5-2.5, elevated 2.5-3.5, euphoria above 3.5. It also carries mvrv_7d_change_pct, active_addresses, issuance_usd_24h and a chain-wide exchange_net_flow_usd_24h.
curl "https://cryptodataapi.com/api/v1/on-chain/dormancy/btc" \
-H "X-API-Key: cdk_live_your_key"The gotcha, and the reason metric_dates exists: Coin Metrics opens a day's row before MVRV lands in it. Until 2026-10-17 that made mvrv read null for part of every day. The endpoint now reads each metric from the newest row that has it and reports the day each value is from in metric_dates, so MVRV and the flows can legitimately trail as_of_date by one day. Treat a one-day gap as normal, not stale.
The deeper treatment of the zones as an agent's macro bias is in MVRV Zones as Cycle Position; the indicator-by-indicator tour is in 11 Bitcoin Cycle Indicators via API. This post is only about the switch from Glassnode.
Glassnode vs This Endpoint: An Honest Comparison
Prices and plan facts below are from Glassnode's own pricing page as of October 2026 and can change; everything on our side is from the live docs.
| Glassnode (Oct 2026) | cryptodataapi.com | |
|---|---|---|
| API access | Professional, from $584/mo billed annually, API as an add-on | Free tier, any cdk_live_ key |
| Cheapest plan with MVRV | Advanced, $49/mo billed annually, Studio only, no API, personal use | Free: 100 requests/day, 1,000 after email verification |
| Resolution | Daily on Advanced; up to 10-minute on Professional | Daily only |
| Metric breadth | Hundreds of metrics, entity-adjusted, multi-asset | Ten BTC cycle indicators + MVRV zones |
| History | Full history per metric | Daily from 2010-2014 per indicator (days=0) |
| Source | Glassnode's own node pipeline | Computed from Coin Metrics' public daily history |
| Rainbow / bubble / peak charts | n/a (CoinGlass-style series) | Not served; 410 with a pointer to the archive |
Where this wins: cost, a key an agent can mint itself, and shapes that join onto the rest of the API (funding, regimes, liquidations) with one header. Where Glassnode wins: breadth, intraday resolution, and cohort-level metrics we do not compute. Pick by what your strategy actually reads.
A Daily Cycle Gate in Ten Lines of Python
Most bots use these indicators as a position-size multiplier, not a trade trigger. Here is the whole job: pull the latest MVRV Z-Score and Puell, derive a 0-1 exposure cap, and let the intraday logic size inside it.
import httpx
BASE = "https://cryptodataapi.com/api/v1"
H = {"X-API-Key": "cdk_live_your_key"}
ci = httpx.get(f"{BASE}/market-intelligence/btc/cycle-indicators",
params={"days": 1}, headers=H, timeout=30).json()["indicators"]
z = ci["mvrv_zscore"][-1]["mvrv_zscore"]
puell = ci["puell_multiple"][-1]["puell_multiple"]
if z > 7 or puell > 4: # historical top territory
exposure_cap = 0.25
elif z < 0 or puell < 0.5: # historical bottom territory
exposure_cap = 1.0
else:
exposure_cap = 0.6
print(z, puell, exposure_cap)Thresholds above are the textbook ones; calibrate them on the days=0 history before trusting them. One pull per day is one request, so even the unverified 100/day free allowance leaves room for everything else the bot reads.
Over MCP the same data arrives as a tool call: claude mcp add --transport http cryptodataapi https://cryptodataapi.com/mcp connects keyless with a browser sign-in on the first call.
When to Use This (and When Not To)
Use it when your strategy gates on daily BTC cycle position: MVRV Z-Score, NUPL, Puell, Pi Cycle, the 200-week MA, AHR999. One free call a day covers all of it with full history for calibration.
- Daily, not intraday. The newest point is usually yesterday. Do not poll it every minute expecting change.
- BTC only. There is no ETH MVRV here. Glassnode and Coin Metrics' paid tiers cover other assets.
- Ten indicators, not hundreds. No SOPR, no cohort HODL waves, no exchange-entity balances per fund.
- Three retired names return 410.
rainbow_chart,bitcoin_bubble,bull_market_peakhave only their pre-2026-07-22 CoinGlass history, in/backtesting/snapshots?data_type=coinglass_cycle_indicators(Pro Plus). - Expect a one-day MVRV lag on the dormancy endpoint and read
metric_datesrather than assuming everything is as ofas_of_date.
Pair the cycle read with a faster signal for timing: the hash ribbon for miner capitulation, or the derivatives layer for crowding. Cycle indicators tell you how much to hold, not when to click.
Frequently asked questions
Is there a free alternative to the Glassnode API for MVRV?
Yes. GET /api/v1/market-intelligence/btc/cycle-indicators returns MVRV Z-Score, NUPL, Puell Multiple and seven other BTC cycle indicators as daily series on the free tier, computed from Coin Metrics' public history. As of October 2026 Glassnode's API requires its Professional plan, from $584 a month billed annually.
How far back does the cycle indicator history go?
Daily from 2010-2014 depending on the indicator. Pass days=0 to get the whole series (about 8 MB, so request gzip); the default is the last 30 entries per indicator.
Why does the rainbow chart return 410?
rainbow_chart, bitcoin_bubble and bull_market_peak were CoinGlass-only series with no free source. Their history up to 2026-07-22 is archived in /backtesting/snapshots?data_type=coinglass_cycle_indicators; the live endpoint returns 410 Gone for those names.
Why is MVRV sometimes a day behind on /on-chain/dormancy/btc?
Coin Metrics opens a day's row before MVRV is in it. The endpoint reads each metric from the newest row that has it and reports that day in metric_dates, so MVRV can trail as_of_date by one day. That is normal, not stale data.



