What 14 days of Hyperliquid builder fees show

The observed totals

• 5,422,505.19 USDC in gross builder fees over 14 UTC days.
• 387,321.8 USDC per day on average across the sample.
• 367–498 builder addresses with builder-coded fills on each day.

These are fees before operating costs, trader rebates, affiliate payouts and platform revenue shares. Non-USDC fees are excluded. We do not convert USDC to dollars or assume every address represents a separate business.

Download the daily data (JSON). Published by @StaxStallion for Floatout on September 25, 2026.

Daily observations

How we measured it

On September 24, 2026, Floatout scanned archived Hyperliquid node_fills_by_block data. Each daily scan read 24 hourly files and hour zero of the following day, then assigned fills to a UTC date using the fill timestamp. Duplicate block numbers were skipped. Builder fill records were deduplicated by trade ID and user.

We counted the distinct builder addresses on the retained records. For fees, we selected only product totals with feeToken = USDC and summed those totals. The archived product totals were rounded to cents before summation, so this report is not an accounting ledger.

Fees are counted per participant fill. If both participants pay a builder fee, both fees belong in the total. That does not mean an execution should count twice as trading volume. We deliberately do not publish a volume or market-share figure here.

The download includes daily product subtotals, block and fill counts, skipped duplicates, and SHA-256 hashes of the original daily scan outputs. Those hashes identify the inputs used, but are not independent proof that the archive is complete.

For protocol definitions, see Hyperliquid’s builder-code documentation, node data schemas and historical data documentation. Floatout produced the observations on this page. Hyperliquid has not endorsed this study.

What this sample cannot tell you

• It does not establish complete protocol-wide coverage. Reading all hourly files does not prove there were no gaps inside them.
• It does not show net profit, withdrawals, customer acquisition costs or how fees are split after collection.
• It does not identify businesses behind addresses or distinguish independent traders from related activity.
• It does not establish typical earnings for a new platform. Daily address counts include zero-fee activity and must not be used as a count of profitable businesses.

What it means for a community operator

The sample shows an active market for interfaces that route trades. It does not show that adding a logo creates demand. Before paying for earnings activation, test whether your members return and use your platform.

Start with the community pilot guide, compare building versus a ready-made platform, and use current pricing and the earnings calculator with your own observed activity. Floatout sells frontend software, so this report has a commercial context.