Stitch

Google Labs’ AI interface-design tool for exploring UI and frontend code from text or image inputs. Models, export and collaboration features in its 2025 announcement need to be checked against the current entry point.

Scope described by official sources

• The 2025-05-20 official announcement describes text-to-UI, image/wireframe references and design variants
• That announcement describes themes, chat iteration, paste-to-Figma and frontend-code export; these are not entitlements promised to every current account
• The Gemini CLI catalog lists a Stitch extension; listing does not establish account setup, executable access or matching web entitlements

Suggested starting workflow

• Define page goals, content hierarchy and device size, then choose references you own or may use
• Check generation, editing and export in the current interface, recording conditions and versions
• Review code, accessibility and asset rights before using outputs as a prototype; exported UI is not a complete business application

These are editorial starting suggestions, not a workflow tested by this site.

The current documentation entry is reachable but only exposes a Stitch title; this does not establish review of complete operating documentation or privacy terms.

Before you use it

• The 2025 announcement uses Gemini 2.5 Pro. Models, authentication and support differ across extensions, SDK and web; historical announcements do not certify the latest model
• The extension catalog includes third-party projects. Repository/package names do not replace formal support or entitlement evidence; no installation, API enablement or authentication was performed
• Documentation and privacy entries only exposed readable titles, not complete policy text. Price, ongoing allowance, regions, commercial rights and asset-storage lifetime remain unknown
• No images were uploaded and no pages generated; detail and classification do not certify task outcomes