v0.20.3

Ollama, the open-source platform that has become a cornerstone for developers running AI models locally, has rolled out version 0.20.3. This incremental but significant release targets core functionality improvements, most notably for Google's recently launched Gemma 4 model. The update specifically enhances Gemma 4's 'tool calling' feature, which is essential for building AI agents that can execute functions, call APIs, and interact with external systems. This makes the lightweight Gemma 4 more viable for complex, multi-step tasks directly on a developer's machine.

Beyond model-specific upgrades, the release expands the accessible model library within the Ollama App, ensuring users have immediate access to the newest community and official releases. It also addresses bugs related to launching the text-based interface for OpenClaw, improving the stability of the terminal experience. For a platform with over 168k GitHub stars, these refinements are crucial for maintaining a smooth user experience as local AI development accelerates. The positive reaction from the community, with dozens of developers engaging with the release notes, underscores its practical value.