Models · Updated 20 Oct, 11:41 pm IST
OpenAI says GPT-6.1 Sol greatly improves legacy code migration
Why it matters for readers: It shows how AI is getting better at tasks that involve older, less-documented programming languages used in critical systems.
- OpenAI created MultiLangBench to test building a SQLite-compatible engine across 20 programming languages.1
- GPT-6 Astra reached an average test pass rate of 87.4%, up from 26.8% for GPT-5.6 Sol on the benchmark.1
- OpenAI reports big language-specific improvements, for example COBOL from 10.9% to 72.1% and Fortran from 21.4% to 83.5%.1
- OpenAI says GPT-6.1 Sol achieves high accuracy migrating COBOL programs to other languages without changing behaviour.1
Get a brief like this every morning
Uzha reads hundreds of sources and gives you the stories that matter for your work, with every source linked. Free.
Get started