OpenAI Enhances AI Models to Crack Niche Programming Languages, COBOL Migration Accuracy Soars to 97.9%
OpenAI reported on October 20 that enhanced models led by GPT-6 Astra can build and migrate low-resource languages such as COBOL, Fortran and Visual Basic, which still underpin banking, ATMs, weather forecasting and nuclear safety analysis. In MultiLangBench, GPT-6 Astra averaged an 87.4% pass rate across 20 languages, up from 26.8% for GPT-5.6 Sol, while COBOL migration accuracy reached 86.6% at low reasoning effort and 97.9% at higher effort, near the benchmark ceiling. Why it matters: enterprises sitting on decades-old legacy systems may finally modernize them at scale, though OpenAI cautions production migrations still need expert review.
SOURCES
- IT之家2026-10-20 23:45