Rubra is a versatile, open-source tool tailored for local development of AI assistants using a large language model (LLM). Designed with developers in mind, Rubra provides the flexibility and intelligence akin to working with OpenAI's ChatGPT, but in a private, cost-effective environment without the need for API tokens. This tool simplifies the creation of AI-powered applications locally, integrating fully configured open-source LLMs to streamline the development of modern AI agents. Developers benefit from a user-friendly chat UI to interact seamlessly with their models and assistants. Unlike other inference engines, Rubra implements an Assistants API compatible with OpenAI, enhancing efficiency with an optimized LLM. Privacy is a cornerstone, as all operations occur locally, ensuring sensitive data and chat histories remain secure on the user's machine. Moreover, Rubra extends beyond its local capabilities to support OpenAI and Anthropic models, fostering community engagement through discussions, bug reporting, and contributions on its GitHub repository.