Recipe 15 of 32
Is smart money exiting token X?
Screen Ethereum tokens with smart_money_only=true, find token X in the list and read net_flow_usd; negative over 7 and 30 days means the cohort sent out more of it than it received.
GET /v1/onchain/tokens/screener?smart_money_only=true counts only the flows of the smart-money cohort, Ethereum
wallets with more than $100,000 of realized profit by our figures. Each token row has the cohort's
in_usd, out_usd and net_flow_usd for the window. The screener has no token filter, so take the
widest list, 100 tokens by cohort volume, and pick token X out of it. The example reads PEPE. The full parameter list is on Token screener.
Quick answer
# smart-money flow in PEPE over 7 days, picked out of the top 100 tokens by cohort volume
curl -s -H "X-API-Key: $VIKE_API_KEY" \
"https://data.vike.io/v1/onchain/tokens/screener?smart_money_only=true&sort=volume&window=7d&limit=100" \
| jq '.[] | select(.token_address == "0x6982508145454ce325ddbe47a25d4ec3d2311933") | {token_address, symbol, net_flow_usd, volume_usd, in_usd, out_usd, unique_holders}'Example response
Captured from the live API on 2026-10-24 at 17:16 UTC, the PEPE row of 100.
Over 7 days the cohort's 319 active wallets sent out $1.59M more PEPE than they received, on $613M
of volume. The same call with window=30d returned net_flow_usd -54,451,561.98 on $1.71B across
429 wallets. Both windows are negative: a steady net outflow, about 3% of the cohort's 30-day
volume. USD values are plain dollars.
{
"token_address": "0x6982508145454ce325ddbe47a25d4ec3d2311933",
"symbol": "PEPE",
"net_flow_usd": -1589015.7,
"volume_usd": 612695003.23,
"in_usd": 305552993.76,
"out_usd": 307142009.47,
"unique_holders": 319
}Variations
The longer window.
# the same row over 30 days
curl -s -H "X-API-Key: $VIKE_API_KEY" \
"https://data.vike.io/v1/onchain/tokens/screener?smart_money_only=true&sort=volume&window=30d&limit=100" \
| jq '.[] | select(.token_address == "0x6982508145454ce325ddbe47a25d4ec3d2311933") | {token_address, symbol, net_flow_usd, volume_usd, in_usd, out_usd, unique_holders}'Every token the cohort is leaving. List the screened tokens with a negative cohort net flow, largest outflow first.
# tokens the cohort sent out on net over 7 days
curl -s -H "X-API-Key: $VIKE_API_KEY" \
"https://data.vike.io/v1/onchain/tokens/screener?smart_money_only=true&sort=volume&window=7d&limit=100" \
| jq '[.[] | select(.net_flow_usd < 0) | {symbol, token_address, net_flow_usd, volume_usd}] | sort_by(.net_flow_usd)'Where it is flowing in. sort=smart_inflow ranks tokens by the cohort's net inflow, largest first.
# the tokens the cohort bought most on net over 7 days
curl -H "X-API-Key: $VIKE_API_KEY" \
"https://data.vike.io/v1/onchain/tokens/screener?sort=smart_inflow&window=7d&limit=20"Before you rely on it
Ethereum only, and only if the token is in the top 100
The cohort exists on Ethereum only: smart_money_only or sort=smart_inflow on another chain is a
400. There is no token filter and no ascending order, so a token outside the cohort's top 100 by
volume cannot be read at all. Net flow counts transfers, not trades: a deposit to an exchange is an
outflow, and a transfer between two cohort wallets cancels out. unique_holders counts cohort
wallets that moved the token, not holders. A window is counted in calendar dates, so 7d covers
today and the seven dates before it.
The cohort rests on a realized-PnL figure with known errors
Membership is decided by the same stored realized PnL that the wallet leaderboard warns about, where a sample found many values far too large. Some wallets in the cohort may not be profitable traders at all. Read the result as what a broad set of large, active wallets is doing.
Ask your agent
Connected to the Vike MCP server (https://vike.io/mcp) in Claude, ChatGPT or Cursor, ask it in plain words.
The token_screener tool answers it, at 3 credits a call.
Is smart money selling PEPE on Ethereum? Compare the last 7 and 30 days.