DEVONthink 4.1 adds generative AI integration, advanced scripting, math support

DEVONthink has released the first feature release for DEVONthink 4.1 Europa, introducing expanded generative AI integration with support for a wider range of server-based models via OpenRouter or any OpenAI-compatible provider, including Amazon, Meta, Microsoft, xAI, Baidu, Moonshot AI, and Qwen. It also adds native support for Claude 4.1 Opus, Gemini 2.5 Flash Lite, OpenAI GPT-5, Google Imagen 4, and the latest Mistral models, plus a new option to connect institutional Ollama installations.

The update brings an AI-powered Script Assistant that generates AppleScript and JavaScript for Automation with automatic error correction, along with expanded AI image editing for both generated and imported images. Mathematical functions and constants are now calculated locally, and academic users benefit from direct retrieval-augmented generation access to Cornell University’s arXiv repository. On macOS Sequoia and later, a new Quick Look extension enables previewing of Spotlight results.

Additional enhancements include faster editing of large Markdown documents, improved performance when updating and sorting lists, and a quicker process for emptying the application’s trash.