THE DOSSIER · ALL ENTITIES

Frontier-model coding marketing built on SWE-bench Verified

The permanent record for Frontier-model coding marketing built on SWE-bench Verified. Every claim we have put on the ledger, checked against independent sources, with the verdict and the date. On the record so far, Frontier-model coding marketing built on SWE-bench Verified is on the record, but the sample is too small to call a pattern. This page compounds: each new verdict updates the rate.

40BS RATE / 100

of the 1 claim we actually checked1

0 VERIFIED 1 TRUE, BUT 0 BS

MOST RECENT ON THE RECORDThe coding number everyone quotes says AI has nearly solved software engineering: Claude Opus 5 scores 96% on SWE-bench Verified. Move to the benchmark built to resist contamination and the top model sits at 80.3%, and GPT-5.6 Sol lands at 64.6%.TRUE, BUT 19 Aug 2026 · 6 SOURCES

THE FULL RECORD · 1 CLAIMS, NEWEST FIRST

TRUE, BUTThe coding number everyone quotes says AI has nearly solved software engineering: Claude Opus 5 scores 96% on SWE-bench Verified. Move to the benchmark built to resist contamination and the top model sits at 80.3%, and GPT-5.6 Sol lands at 64.6%.CAPABILITY · 19 Aug 2026 · 6 SOURCES

SEARCH EVERY FRONTIER-MODEL CODING MARKETING BUILT ON SWE-BENCH VERIFIED CLAIM IN THE LEDGER →
← ALL ENTITY DOSSIERS