Hi, this is Steve Ruiz. You might know me from tldraw, a free and open source collaborative whiteboard that I released last year. That project got quite popular: users liked the simplicity and attention to detail, while developers saw its React-based “canvas” as the perfect starting point for their own adventures in spatial user interfaces.

This last year, we’ve been hard at work on bringing that idea forward.

The new tldraw (which you can try here) is a full rewrite, built from the ground up to provide a primitive for spatial user interfaces that can display native web content on an infinite canvas. Like the original, tldraw is both an app and a library, designed for integration and extension. The new project goes further, introducing a custom engine for real-time collaboration.

Today I’m announcing tldraw’s $2.7 million seed round led by Lux Capital and including the support of Amplify Partners, Sabrina Hahn, Guillermo Rauch of Vercel, Liu Jiang, NP-Hard Ventures, Soleio, Badrul Farooqi, Michael Stoppelman, Tom Preston-Werner, Adam Wiggins of Muse, Brian Lovin of Campsite, David Khourshid of Stately.ai, Cristóbal Valenzuela of Runway ML, Johannes Schickling of Prisma, Brian Shultz of Tango, Shin Kim of Eraser, Jeff Weinstein and Chase Davis.

Subscribe for updates or read on for more!

Where it’s been

From the beginning, tldraw was more than an app: it’s distributed as a React component that can be embedded in other apps; and, being open source, it’s been free to fork and modify to create new experiences.

Developers have used tldraw to build some truly amazing things.

A surprise to me was that, with a few notable exceptions (such as the fantastic okso.app and Jordan Singer’s Macpaint app), most of these new projects were less about drawing or white-boarding and more about putting interactive widgets into a Figma-like interface.

Affine used tldraw to create their “edgeless” view of their Notion-style blocks. Legend Keeper and WorldAnvil both use tldraw to include characters and places from users’ story-worlds onto the canvas. Vidext is using tldraw to create AI-driven videos, BigBlueButton has reimplemented their virtual classroom’s whiteboard with tldraw, and Logseq is building their whiteboards feature on tldraw, too.

However, as I spoke with developers about what they were making with tldraw, a few things became obvious:

1. that the community needed an open-source primitive for spatial interfaces;

2. the canvas was only part of the question; and

3. what was needed was much bigger than what I’d made.

And so we started making something new.

Where it’s going

The new tldraw is designed to be a primitive for infinite canvas applications, providing the same type of infrastructure utility that ProseMirror provides for rich text editors or Mapbox provides for maps. Like text editors and maps, a canvas is a nightmare of internal complexity, both technical and in user experience design, together with a long list of table-stakes features that need to accompany any product.

It’s our belief that a canvas should be a thing you build with, rather than build yourself.

And so the new tldraw library comes batteries included—much more-so than the original tldraw library ever did—with solid implementations for complex features such as selection, rotation, translation, resizing, styling, erasing, camera controls, editable text, reordering, snapping and alignment, support for multiple pages, undo and redo, copy and paste, embedded content, image export, file persistence, digital ink, and of course perfect arrows.

And, just like maps and text editors, the story doesn’t end there.

Our new canvas is also joined by our new high-performance service for collaboration, providing an easy path to creating multiplayer experiences. You can try this on our beta site by visiting beta.tldraw.com/r and sharing the URL with up to 30 friends.

Like the core library, we intend to make this service available for teams to build with. And we’ll be rolling out new services next year for related features: comments and messages, history and versioning, permissions and all of the other bits and pieces that go into making a spatial collaboration experience.

What’s next

We’ve been building the team: I’m joined by David Sheldrick, Jaime Blair, Lu Wilson, and Alex Dytrych, together with Faraz Shaikh, Mitja Bezenšek, and Sunil Pai. We have a lovely office in London (Thursdays are for visitors) and of course we’re hiring.

While we haven’t open-sourced the new tldraw yet, we are looking for design partners to help get us there. If you are interested in getting early access to the new tldraw library, please let me know at steve@tldraw.com or schedule a call.

Until then, please give the new tldraw a spin at beta.tldraw.com. Share some projects! Let us know on Twitter or Discord if you have any questions, feedback, issues, or ideas.

P.S. A very special thanks to the hundreds of individual and corporate sponsors that helped me bring this project to where it is today. This would not have happened without your support. Thank you!