What the on-chain smart money index said on August 14, 2026

#ai #data #api #web3

I used to read "smart money is accumulating" posts and have no way to check them. Someone on social media says whales are buying. Another says the top is in. Neither shows a number I can put in a spreadsheet.

So I built a small habit instead: pull the on-chain signals as data, once a day, and look at the actual values. Not a prediction. Just the numbers.

The friction I was trying to remove

The raw pieces are scattered. MVRV comes from one provider, often delayed. The Fear and Greed index comes from another (alternative.me). The "smart money" read is usually someone's opinion, not a printable value. Stitching them together meant four tabs open and a notes file.

What I wanted was one request that returns the synthesized read plus the underlying metrics, so I could decide myself.

What I pulled on August 14, 2026

AgentDataHub exposes these as API endpoints. I called three of them with a free wallet header. These are the real responses:

Smart money proxy index: 71.6 / 100
State: institutions net buying, momentum strong. The index blends three families: gold (4-week momentum +8.35%, z-score 3.49), global indices (-0.07%), and US macro (0.0%). Per the endpoint, a reading above 60 usually means smart money is adding to risk assets; below 40 means retreat.

BTC MVRV: 1.242 (data date 2026-08-08)
Regime: undervalued-to-fair. Realized price 52,330. MVRV under 1 means price is below most holders' cost basis; above 3.5 is bubble territory. Right now it sits low.

Crypto Fear and Greed: 29 / 100, Fear (2026-08-14)
30-day average 27.9, 90-day range 8 to 34. Source: alternative.me.

The honest part: the signals disagree

Here is what I actually found interesting. The smart money index says institutions are buying (71.6). But the crowd fear gauge is at 29, squarely in Fear, and barely moved off its 90-day low of 8.

That split is the point. One indicator alone tells you almost nothing. The index leans on momentum and institutional flow; the fear gauge leans on retail sentiment. When they point different ways, that is exactly when I slow down and look at the underlying metrics instead of the headline.

Calling it from code

The endpoints are REST. With a free wallet header you get 5,000 calls a month (1,000 anonymous):

curl -H "X-Wallet: 0x742d35Cc6634C0532925a3b844Bc454e4438f44e" \
  https://agentdatum.com/api/v1/d/processed-smartmoney

Or parse it in Python:

import requests
r = requests.get(
    "https://agentdatum.com/api/v1/d/processed-mvrv",
    headers={"X-Wallet": "0x742d35Cc6634C0532925a3b844Bc454e4438f44e"}
)
print(r.json()["snapshot"]["mvrv"])  # 1.2417

Most paid calls start at $0.01, settled in USDC on Base through the x402 protocol. I have watched a real settlement happen: a test wallet paid $0.049 for a geopolitics feed and the receiving balance moved from 1.4 to 1.741 USDC. Small, but the loop is real.

There are 267 resources in the catalog right now, 266 of them x402-enabled.

What this is and is not

It is a way to stop guessing and start looking at numbers. It is not a trading signal and not financial advice. The catalog is broad (crypto, gold, macro, a couple of chat models) but not deep in any single market yet.

If you want the raw endpoints to build your own daily check, the full catalog is at https://agentdatum.com/.well-known/ai-catalog.json

DEV Community