Zeplin AI Annotator suggests design tokens and performs automated accessibility checks

Design · 4 min read

When designers publish components to Zeplin, AI Annotator analyzes properties like spacing, color, and typography and proposes tokens consistent with existing token sets. It flags accessibility issues—contrast failures, missing focus states, and ambiguous hit targets—and offers code snippets that developers can use to remediate problems quickly.

The tool integrates with design systems and CI pipelines so suggested tokens can be approved and merged into a team's centralized library. Zeplin described the feature as a way to close the loop faster between visual design and implementation, with fewer ambiguous handoffs.

Teams testing the feature reported faster iteration cycles and fewer accessibility regressions. Still, some designers want finer-grained control over autogenerated token names and defaults to avoid token proliferation across large organizations.