Editing Portal:Artificial Intelligence (section)

=== PerlGPT ===

'''PerlGPT''' is one of the most visible Perl-specific LLM efforts. The Perl and Raku Foundation published a 2023 grant proposal for “PerlGPT Large Language Model, Phase 1.” The proposal described PerlGPT as a large language model trained on additional Perl-related content only, based on Meta’s Code Llama models, with new components implemented in Perl where possible and released as free and open-source software.<ref>The Perl and Raku Foundation, “PerlGPT Large Language Model, Phase 1,” https://news.perlfoundation.org/post/perlgptphase1, accessed July 3, 2026.</ref>

The proposal described Phase 1 as training a 13B input language model using Perl-related stimulus/response pairs curated from PerlMonks, Stack Overflow, GitLab, GitHub, and other public Perl-specific data sources, with the goal of generating Perl source code in collaboration with Perl programmers.<ref>The Perl and Raku Foundation, “PerlGPT Large Language Model, Phase 1,” https://news.perlfoundation.org/post/perlgptphase1, accessed July 3, 2026.</ref>

PerlGPT is historically important even apart from its eventual technical outcome. It represents a recognition that general-purpose coding models often underrepresent Perl compared with Python, JavaScript, Java, C++, and other more heavily represented languages. A Perl-focused model or curated Perl training set could help preserve Perl idioms, CPAN knowledge, documentation patterns, and maintenance practices.