AIThe Decoderabout 4 hours ago
AI model ran 19 days on $2,600 coding task
1 min read
Epoch AI and METR released MirrorCode, a benchmark where AI models reimplement programs from scratch without source code. Claude Opus 4.7 leads with 56 percent solve rate. One task cost $2,600 and ran continuously for 19 days with no human involvement. Opus 4.7 reimplemented a 16,000-line bioinformatics toolkit in 14 hours for $251.
Level
Hype check
Tap to vote and see what everyone thinks.
