OpenAI announces improved COBOL building and migration capabilities
On October 19, 2026 OpenAI said it has enhanced its ability to generate new COBOL programs and to migrate existing COBOL code to other languages. The company shared results from two internal benchmarks. In MultiLangBench, which tests building a SQLite engine from scratch, the average pass rate across 20 languages was 87.4% using the GPT‑6 Astra model with maximum inference settings. For COBOL…
Key points
- OpenAI reported an 87.4% pass rate on MultiLangBench for 20 languages using GPT‑6 Astra max inference.
- GPT‑6.1 Sol achieved 86.6% accuracy on low‑inference COBOL migration and 97.9% on high‑inference.
- OpenAI cautions that migration still needs expert review and may not preserve all business rules.
OpenAI emphasized that these scores do not guarantee flawless migration of whole business systems. It warned that expert review and testing against specific business rules are still required, as subtle differences in calculations can change outcomes. The announcement is framed as an evaluation of existing models rather than the launch of a COBOL‑specific AI product.
Model pages: GPT-6 Astra → · GPT-6.1 Sol →
The story so far
3 episodes →- OpenAI announces improved COBOL building and migration capabilitiesthis story
[Breaking News] OpenAI's COBOL support evolves. The 'old work rules' that AI will take over
note.com · 20 October 2026
Loading the full article…
This text was published by note.com and written by ダニエル / DADION. It is reproduced here with attribution so you can read it in full; the rights remain with the publisher. Read it at the source ↗
The headline, key points and digest above were generated by Digest AI's editorial model from the linked sources. Automated summaries can contain errors: the sources are the record. Spotted a mistake? Tell us. Published by Martin K., who runs Digest AI and handles corrections.
More in Research
All →- OpenAI's GPT-5.6 Sol and Google's Gemini 4 Argon find critical PHP flaw · 1 src
- PHP discloses PHP2Shell vulnerability affecting up to 70% of websites · 8 src
- Ggml-org adds fused top‑k mask operation to llama.cpp · 1 src
- Microsoft Research says Xinzhi Zhang and Zhao Song prove KLS conjecture with AI · 1 src
- Demis Hassabis says he expects AGI within two to three years, down from his three-to-five-year prediction earlier in 2026, and discusses AI's science potential · 2 src
Comments
via GitHub Discussions