The new OpenUI platform enables large language models to create interactive user interfaces on the fly, reportedly reducing token consumption by roughly two‑thirds.
Thesys has launched OpenUI, a framework that lets large language models generate full‑featured dashboards and forms in real time, cutting token usage by roughly two‑thirds compared with previous generative UI approaches.
Why OpenUI matters for AI‑driven interfaces
Traditional AI‑generated interfaces require multiple prompt‑completion cycles, each consuming tokens that quickly add up. OpenUI consolidates layout, data binding, and interaction logic into a single, reusable schema, allowing the model to produce a complete UI with far fewer calls.
Core features of the OpenUI platform
- Declarative component library covering charts, tables, forms, and navigation elements
- Built‑in token‑optimisation engine that merges similar UI elements before rendering
- Automatic accessibility compliance checks
- Extensible plug‑in system for custom widgets and data sources
Developers can embed OpenUI into existing AI agents via a lightweight SDK, enabling on‑the‑fly UI generation for tasks such as data analysis, report creation, and workflow automation.
Potential impact on product design workflows
By standardising the output format of AI‑generated interfaces, OpenUI promises smoother handoffs between AI prototyping tools and human designers, reducing the need for manual re‑creation of layouts.
For readers interested in broader trends in AI‑assisted design, see the Product Design section on uxc.news.
Read the full announcement in The NextGen Tech Insider coverage of Thesys OpenUI launch.
Track coverage on Momentum · More in Product & Design.