Equip your product team with the power of AI
Custom pricing tailored to your team's needs
Save and reuse your projects as templates across your team
Create and manage branches for your projects with full version history
Bring your real components in Onlook and use them in your projects
Centralized design tokens, color palettes, and typography management
Navigate your React component tree with precise control over every element
Get instant help and generate code with unlimited AI-powered assistance
Built with the community. Customize and extend for your team's needs
Deploy your projects to your own internal domain
SSO (SAML/OAuth), advanced security controls, audit logs, and admin controls
Existing paid plan users can continue using Onlook. New users – Please contact us or book a demo to get your team set up. If you're looking to self-host Onlook, please check out the GitHub repository or reach out to us to schedule a call.
Onlook is a visual design canvas that connects to your existing codebase. Designers drag real components onto an infinite canvas, make changes visually, and submit pull requests — no coding required.
Traditional design tools create static mockups that must be rebuilt in code. Onlook works with your real components — what you design IS the code. Changes become PRs, not handoff specs.
AI generators create new code from scratch. Onlook constrains AI to YOUR existing components, so outputs match your design system. No translation, no drift.
No. Designers use a visual canvas with familiar tools. Real code runs underneath — you don't need to touch it unless you want to.
Yes. Share your canvas, leave spatial comments, and work together in real-time. Changes sync to code and can be submitted as PRs for engineers to review.
React, Next.js, and any CSS approach (Tailwind, CSS modules, styled-components). Works with any component library.
Yes, Onlook can be self-hosted for free on GitHub. For the hosted cloud version, please contact our team or book a demo.
The code you make with Onlook is all yours. Export it locally, publish to GitHub, or deploy to a custom domain.