ollama4j

Why we included this project

ollama4j is a plain Java wrapper around the Ollama server, so teams on the JVM can reach local language models without hand-rolling HTTP calls. It exposes typed methods for pulling models, running single-pass or multi-turn chats, streaming token output, and requesting embeddings, with built-in support for image inputs, tool or function calling, and MCP-based external tools. Because you run the Ollama server yourself, model traffic stays on your own hardware or a private endpoint while the rest of the application is ordinary Java. That makes it a practical choice for adding generative features to backend JVM services, where the alternative is maintaining REST glue against Ollama's raw API by hand. A type-safe options builder keeps request configuration readable, and hooks for timeouts, authentication, and logging leave the integration code tidy and easy to test.