Hugging Face, StarCoder (open-source)

StarCoder and StarCoderBase are Large Language Models for Code (Code LLMs) trained on permissively licensed data from GitHub, including 80+ programming languages, Git commits, GitHub issues, and Jupyter notebooks. Article, paper, code.

Google DeepMind, AlphaCode (not public, paper and dataset are available)

AlphaCode is a system for code generation that can create novel solutions to problems that require deeper reasoning. In simulated evaluations on the Codeforces platform, AlphaCode achieved, on average, a ranking of the top 54.3% in competitions with more than 5,000 participants. Article, paper, dataset.

Carnegie Mellon University, Polycoder (open-source)

PolyCoder is a 2.7B parameters model based on the GPT-2 architecture, that was trained on 249GB of code across 12 programming languages on a single machine. Paper, code.

Salesforce Codegen, CodeT5 (open-source)

CodeGen is a family of open-source models for program synthesis. Article, papers and code.

CodeT5 can be deployed as an AI-powered coding assistant to boost the productivity of software developers. It includes a demo using CodeT5 as a VS Code plugin to provide three capabilities: text-to-code generation, code autocompletion, and code summarization. Article, papers and code.

Replit, replit-code model (open-source)

replit-code-v1-3b is a 2.7B Causal Language Model focused on Code Completion. The training mixture includes 20 different languages, listed here in descending order of number of tokens: Markdown, Java, JavaScript, Python, TypeScript, PHP, SQL, JSX, reStructuredText, Rust, C, CSS, Go, C++, HTML, Vue, Ruby, Jupyter Notebook, R, Shell. In total, the training dataset contains 175B tokens.

Other open-source LLMs

There are many other open-source LLMs that are not specialized in coding but achieve significant results in this area. You can learn about some of them in our article about top open-source LLMs.

AWS, CodeWhisperer (free for individual use)

CodeWhisperer is a coding assistant that provides developers with real-time code suggestions directly in the integrated development environment (IDE). Website

Replit (free plan included)

Replit allows you to write code with Ghostwriter, an AI-powered pair programmer, collaborate in their real-time multiplayer editor, and build, test, and deploy directly from the browser. Website

Tabnine (free plan included)

Tabnine is an AI coding assistant that autocompletes lines of code, suggests full function completion based on the function declaration, and generates blocks of code based on natural language comments. Website

Sourcegraph, Cody (free plan included)

Cody answers technical questions and writes code directly in your IDE, using your code graph for context and accuracy. Website

GitHub, Copilot (paid, $10 for individuals)

GitHub Copilot is an AI pair programmer. It draws context from comments and code to suggest individual lines and whole functions instantly. GitHub Copilot is powered by a generative AI model developed by GitHub, OpenAI, and Microsoft. It is available as an extension for Visual Studio Code, Visual Studio, Neovim, and the JetBrains suite of integrated development environments (IDEs). Website

OpenAI, ChatGPT (free version included)

Google, Bard (free)

Anthropic, Claude (free version included)