New gas cost proposal for EIP-7904

This is an automated report generated from the script ./src/estimate_7904_repricings.py. The script uses the runtime estimation output generated by this script. The report with the runtime estimation results can be found in ./reports/eip-7904/2026-01-15_2026-01-29/runtime_estimation_autogenerated_report.md.

Methodology

New gas costs are calculated using an anchor rate of 60 million gas per second, which represents a target execution rate for EVM operations. The formula used is:

new_gas = (anchor_rate * runtime_ms) / 1000

Where runtime_ms is the estimated runtime in milliseconds from the regression models.

Understanding the results

The table below shows the worst-case gas costs across all tested clients (taking the maximum estimated cost per operation). This conservative approach ensures that the new gas costs account for the slowest implementation among the major Ethereum clients.

The Change column shows the relative change as a decimal (e.g., 1.0 = 100% increase, 0.5 = 50% increase, -0.25 = 25% decrease). Operations with inf indicate costs going from 0 to a positive value.

Only operations and parameters with good model fits (R² > 0.5 and p-value < 0.05) are included in the gas cost proposals. Operations with poor model fits are listed separately in the “Errors and caveats” section.

New gas proposal

The following table shows the new gas cost for all operations and parameters with a good model fit.

Opcode Parameter Current Gas New Gas (Rounded) Change
ADDMOD constant 8 8 0.0
BLAKE2F constant 0 170 inf
BLAKE2F num_rounds 1 2 1.0
BLS12_G1ADD constant 375 643 0.71
BLS12_G2ADD constant 600 765 0.27
DIV constant 5 15 2.0
ECADD constant 150 314 1.09
ECPAIRING constant 45000 34710 -0.23
ECPAIRING num_pairs 34000 34103 0.0
ECRECOVER constant 3000 2904 -0.03
KECCAK256 constant 30 45 0.5
KECCAK256 msg_size 6 1 -0.83
MOD constant 5 12 1.4
MULMOD constant 8 11 0.38
POINT_EVALUATION constant 50000 89363 0.79
SDIV constant 5 20 3.0
SMOD constant 5 3 -0.4

Gas costs by client

The following plot shows the new gas costs (rounded) for each operation parameter across different clients, with error bars representing the confidence intervals.

Gas costs by client

Errors and caveats

All parameters had a good model fit.

The following operations have no estimation for the following clients:

  • BLAKE2F: erigon

  • BLS12_G1ADD: erigon

  • BLS12_G2ADD: erigon

  • ECADD: erigon

  • ECPAIRING: erigon

  • ECRECOVER: erigon

  • POINT_EVALUATION: erigon