Ollama v0.20.5 was released on April 9, 2026, introducing significant new features for local AI model interaction and performance. The update includes the OpenClaw channel setup, which allows users to connect Ollama instances to popular messaging platforms such as WhatsApp, Telegram, and Discord via the ollama launch openclaw command. This aims to broaden the accessibility and integration of local AI capabilities into daily communication workflows. Additionally, Ollama v0.20.5 enables flash attention for Gemma 4 models when running on compatible GPUs, a performance optimization designed to enhance inference speed. The release also improves usability by enabling ollama launch openclaw to detect curl-based OpenCode installations at a common path, and it resolves an issue with the /save command for models utilizing safetensors-based architectures.