EIP-8131
Unified Transaction Content Floor
Prices every user-controlled byte in a transaction at a single flat floor rate, so the worst-case block size is a predictable function of the gas limit.
The problem
Transaction-content pricing grew up piecemeal. Calldata had one rule (EIP-7976, 64 gas/byte), access-list entries another (EIP-7981, fixed surcharges), and newer fields (EIP-7702 authorization tuples, EIP-4844 blob hashes) had no floor at all. Each new transaction field meant another bespoke pricing EIP, and the gaps left room to stuff bytes into a block more cheaply than the protocol intended.
What it does
EIP-8131 collapses all of that into one rule: every user-controlled content byte is priced at a flat 64 gas per byte at the floor.
tx_floor = 21_000
+ 64 × ( calldata_bytes
+ 20 × addresses
+ 32 × storage_keys
+ 108 × auths
+ 32 × blob_hashes )A transaction must satisfy tx.gas ≥ max(intrinsic_gas, tx_floor), and its accounted
usage is gas_used = max(execution_gas, tx_floor). The constants reflect the worst-case
encoded size of each field (e.g. 108 bytes for an authorization tuple, 32 for a blob hash).
Why it matters for scaling
Because every content byte now has a known floor price, the maximum bytes an attacker
can pack into a block becomes pure arithmetic, roughly gas_limit / 64, with no
reliance on compression. At a 60M gas limit that bounds attacker-controlled content to
about 0.89 MB. That predictable ceiling is exactly what the bandwidth side of the
anchor analysis needs to safely raise the limit.
Impact
- Aggregate network gas rises ~3.56%.
- Fewer than 4% of mainnet transactions hit the new floor.
- The most-affected class is EIP-7702 (SetCode) transactions; their bind rate to the floor rises from 0.04% to ~20.6%.
- Type-3 (blob) transactions see their average floor rise from ~69k to ~84k gas.