FluidInference released a compact cua-s1-forms model for iOS

Team FluidInference introduced a specialized classifier cua-s1-forms, optimized for Apple devices, designed for automatic action detection in forms and fully adapted to the Core ML format.

The model size is only 706 thousand parameters, allowing it to be integrated into mobile applications without significantly affecting performance, requiring iOS 17 or later to function correctly.

During testing on a set of 196 test strings, the model demonstrated 100% accuracy in classification, with developers providing ready-to-use examples of implementing interaction logic with the model in Python and Swift, simplifying deployment in existing projects.

Using Core ML enables computations to be performed directly on the device's hardware, eliminating the need to send data to a server, making this solution suitable for tasks requiring high-speed processing of user input in real-time while maintaining data confidentiality.