SCDBench (Compilation Repair) - ABI Recovery: leaderboard

Metric: ABI recovery F1 (%; micro F1 of the recovered public function signatures against the ground-truth ABI over all 600 contracts, extra functions counting as false positives and missing ones as false negatives, printed as a fraction and scaled to percent; after one same-model repair attempt from the compiler errors). Source: arxiv.org. Saturation forecast: Around December 2026. 4 models tracked.

Top models

#ModelScore
1GPT-5.3 Codex (High)89.6
2Claude Opus 4.7 (High)72.1
3GLM-5 (Thinking)69.7
4GLM-5 (Non-reasoning)65.8

Interactive version: theaggregate.ai/benchmark?slug=scdbench-compilation-repair-abi-recovery · How It Works · Data refreshed daily, snapshot 2026-09-29.