JetBrains has announced that its code completion LLM, Mellum, is now available on Hugging Face as an open source model.
According to the company, Mellum is a “focal model,” meaning that it was built purposely for a specific task, rather than trying to be good at everything. “It’s designed to do one thing really well: code completion,” Anton Semenkin, senior product manager at JetBrains, and Michelle Frost, AI advocate at JetBrains, wrote in a blog post.
Focal models tend to be cheaper to run than general larger models, which makes them more accessible to teams that don’t have the resources to be running large models.
“Think of it like T-shaped skills – a concept where a person has a broad understanding across many topics (the horizontal top bar or their breadth of knowledge), but deep expertise in one specific area (the vertical stem or depth). Focal models follow this same idea: they aren’t built to handle everything. Instead, they specialize and excel at a single task where depth truly delivers value,” the authors wrote.
Mellum currently supports code completion for several popular languages: Java, Kotlin, Python, Go, PHP, C, C++, C#, JavaScript, TypeScript, CSS, HTML, Rust, Ruby.
There are plans to grow Mellum into a family of different focal models ideal for other specific coding tasks, such as diff prediction.
The current version of Mellum is most ideal for either AI/ML researchers exploring AI’s role in software development, or AI/ML engineers or educators as a foundation for learning how to build, fine-tune, and adapt domain-specific language models.
“Mellum isn’t a plug-and-play solution. By releasing it on Hugging Face, we are offering researchers, educators, and advanced teams the opportunity to explore how a purpose-built model works under the hood,” the authors wrote.