Today for AI
HOT RADAR
llmHEAT 0.0°

AI CURATED HOT STORY · 10/19/2026

OpenAI GPT-6 Boosts Legacy Language Coding Capabilities

5 reports archived0 independent sourcesupdated 10/21/2026, 15:08:36
Synthesis & Latest Updates

OpenAI has released optimized models for low-resource legacy languages like COBOL and Fortran, with GPT-6 Astra achieving an average pass rate of 87.4% on MultiLangBench, up from 38%. This marks a significant breakthrough in modernizing critical infrastructure.

Cross-verified reports indicate that GPT-6.1 Sol achieves 97.9% accuracy when migrating COBOL to modern languages such as C++ and Java under high-reasoning modes. OpenAI also open-sourced MultiLangBench to evaluate zero-shot coding capabilities across 20 low-resource languages and introduced the cost-effective GPT-6 Luna, reducing API costs by approximately 78% compared to previous generations.

This advancement lowers the technical and economic barriers for refactoring legacy systems, reshaping digital transformation paths for traditional IT assets. By opening API access for developer feedback, AI-assisted maintenance is becoming a new paradigm for ensuring software stability in critical sectors like finance and nuclear safety.

LATEST/OpenAI has officially open-sourced the MultiLangBench evaluation tool and opened API access for developers to test GPT-6 series models. The company emphasizes using real-time user feedback on specific workflows like COBOL and Fortran to create a data loop for continuous model fine-tuning.

Hourly heat curve

168 fully observed hours
Now
0.0
Peak
24.216:00
24h change
–

The line compares only sources observed throughout; gap hours are interpolated to keep the trend continuous.

Coverage timeline

Total 5 reports · Latest first
  1. InfoQ 中文T2·85 pts

    OpenAI GPT-6.1 Sol Boosts COBOL Migration Accuracy to 97.9%

    • GPT-6.1 Sol improved COBOL code migration accuracy from 5.4% to 97.9%, addressing generalization issues caused by scarce training data for low-resource languages.
    • GPT-6 Astra achieved an 87.4% average test pass rate in complex engineering tasks across 20 languages (e.g., implementing a SQLite-compatible engine from scratch), far exceeding the previous model's 26.8%.
    • Despite excellent benchmark performance, translating high accuracy into reliable production migrations still requires comprehensive engineering validation to prove correct execution of business rules.
  2. IT之家 · 智能时代T2·85 pts

    OpenAI Releases GPT-6 Astra with 97.9% COBOL Migration Accuracy

    • GPT-6 Astra achieved 97.9% accuracy in migrating COBOL to C++/Java/Python/Rust, a massive improvement over predecessors.
    • OpenAI introduced MultiLangBench, a new benchmark specifically evaluating zero-shot construction capabilities across 20 low-resource programming languages.
    • The cost-effective GPT-6 Luna variant has an average pass rate of 40.3%, with API costs reduced by approximately 78% compared to the previous generation.