Models

Our approach to model selection is to use the most suitable model for each task, within the constraints of the supported hardware. Currently, Tiles is designed to handle everyday tasks, with plans to expand into more domain-specific use cases over time. We provide a carefully tested combination of system prompts, tools, and models, so users do not need to manage model selection or scaffolding themselves.

We use gpt-oss-20b as the primary model for everyday tasks and expose it through the Open Responses API. Its Mixture of Experts architecture activates only around 4B parameters at inference time out of a total 21B, making it a strong balance between quality and efficiency.

This setup requires 16 GB of unified memory to run optimally and should not pose an adoption issue, as all Macs sold from late 2024 onward ship with 16 GB of base memory.