Skip to main content
EnterpriseModelsTools

OpenAI says GPT-6 is much better with legacy programming languages such as COBOL and Fortran

Source: The Decoder (opens in a new tab) · Matthias Bastian

Intel Summary

According to reporting by The Decoder, OpenAI claims its latest models show substantial improvements on legacy and low-resource programming languages, including COBOL, Fortran, and Visual Basic. On the MultiLangBench evaluation, OpenAI reports GPT-6 Astra achieved an 87.4 percent test pass rate across 20 languages compared to 38 percent for GPT-5.6 Terra, with GPT-6.1 Sol reportedly reaching up to 97.9 percent accuracy on COBOL migration tasks.

Why It Matters

High accuracy in legacy code translation addresses a major operational bottleneck for sectors reliant on aging codebases, particularly financial services and government institutions. If verified in enterprise environments, these capabilities could accelerate long-delayed modernization initiatives and reduce the costs of migrating legacy systems.

Organizations & Entities