Docs

Every number on the page, and the rule behind it.

PlumS shows what a tokenized stock costs across three venues, and what moving to the cheapest one would cost you. This page states how each of those numbers is produced, including the parts that make a reading unreliable.

What a row holds

Each row is one US stock and the four prices being compared: the real US market price, and the same stock tokenized on BNB Chain, on Solana and on Robinhood Chain. Under each venue price sits its premium or discount against the real price. Prices refresh every minute.

Real market
Finnhub, read server-side so the API key never reaches your browser.
BNB bStocks
DexScreener, from a pool pinned per token and quoted in canonical BSC USDT or USDC.
Solana xStocks
DexScreener, from the pinned xStocks pair.
Robinhood Chain
An executable Uniswap quote: one token into USDG across V4, V3, V2 and UniswapX. That is what a trade would actually get, which a pool price is not. DexScreener is the fallback.
Swap routes
LI.FI, which finds the route and executes it.

Premium, discount, and which venue they use

A premium means the token trades above the real share price, a discount below it. The figure that drives sorting, filtering and the headline is read at the cheapest venue that has a usable price, because that is the least anyone can pay above the real price, and it is the same venue the swap panel quotes.

The cheapest venue changes, so the headline names the one it used. A venue whose gap against the real price exceeds 40% is treated as a broken or near-empty pool: it is marked ⚠ thin, it never becomes a destination, and it is left out of every ranking. Without that guard a dead pool would win "biggest discount" every time.

Venue spread

The distance between the cheapest and the dearest venue with a usable price, as a percentage of the cheapest. It is a magnitude, not a direction, so it carries no sign and no colour: a spread has no good or bad side. The bar beside it runs to full scale at 2% and stops there rather than rescaling, so a bar means the same thing on every refresh.

The swap panel

The Swap button on a row asks LI.FI what it would cost to buy that stock at its cheapest venue, for a $100 notional in USDT on BNB Chain. The panel lists the routes it found, what each returns, its gas, and its total cost as a percentage of what went in.

Above the table sits the verdict, which compares that cost against what moving venue actually saves:

Not worth it
The route costs at least as much as the saving.
Marginal
The saving clears the route cost by less than 0.5 points, which is noise.
Worth it
The saving clears the route cost by more than that.
Single venue
Only one venue has a usable price, so there is nothing to compare against.

Most readings come back not worth it. Venue gaps here have run roughly 0.1% to 0.7% while route friction runs 0.3% to 0.6%, so the gap is usually smaller than the cost of capturing it. This is a cheaper-venue convenience, not an arbitrage tool, and the panel is built to say so rather than to sell the trade.

Every figure in the panel is an estimate quoted before any wallet is connected, and prices move between a quote and its execution.

The fee

PlumS adds a 0.2% fee to a route, but only when the saving still beats the route cost plus that fee. When it does not, the fee is zero and the panel says so. A flat cut on every swap would contradict the verdict line on almost every row, so the fee only exists where moving venue genuinely pays.

Whatever fee applies is quoted and executed as the same number: the panel re-quotes with it rather than adding it to the display afterwards, so the cost column is what you actually pay.

What PlumS does not do

PlumS never connects to your funds, never routes an order itself and never takes custody. LI.FI finds and executes routes; the wallet connection belongs to its widget, not to this site. PlumS is not affiliated with Robinhood, Binance, Backed Finance or xStocks, and nothing here is financial advice.

Limits worth knowing

The tokens trade around the clock while the US market does not, so outside trading hours a premium is measured against a price that has stopped moving. Expect gaps to look larger then, and to be less meaningful.

These are permissionless chains, and a symbol search is never evidence: the same query that finds the real SPYB also returns a pool quoting millions a share. Every venue address here was resolved by contract address on an issuer-independent index and then matched to a pool quoted in a canonical stablecoin. Prices can still be wrong, so verify anything you intend to act on.