Launch Video Library · AI · Launch · 2026

Introducing Solaris

Runway introduced Solaris, a new operating system model that generates interactive interfaces in real time without code.

What Runway shipped

Runway continues to push the boundaries of generative AI with the launch of Solaris, an Interface World Model. Rather than generating video pixels for entertainment, this system generates interactive operating system interfaces in real time, frame by frame, without underlying code.

The 83-second launch film reflects the technical weight of the announcement. It relies on stark, confident motion design to explain a fundamentally new paradigm in human-computer interaction, stripping away abstraction to focus on the raw output of the model.

How the motion works

Runway’s motion language is consistently clinical and precise, treating AI research as a hard science. The film opens with stark kinetic typography that establishes the problem space before cutting directly to the product in action. There is no narrative fluff; the edit relies on hard cuts to move between interface generations, mirroring the instant, real-time nature of the Solaris model itself.

To demonstrate the generated UI, the camera work employs a subtle camera drift. Instead of static screen recordings, the viewport slowly pushes in on the generated interfaces. This continuous motion gives a sense of depth to flat UI elements, subtly reinforcing that these are not traditional coded screens, but generated spatial environments.

The pacing accelerates as the film demonstrates the model's versatility across different structural layouts. A sequence of rapid zoom-through transitions pulls the viewer from one generated interface into the next. This creates a recursive, infinite-canvas feel, visually communicating the boundless combinations the world model can produce while maintaining strict information retention.

What to steal from it

  • Use hard cuts instead of crossfades to convey speed and technical precision.
  • Apply a slow camera drift to screen recordings to add spatial depth to flat interfaces.
  • Let the product output drive the visual narrative rather than relying on abstract 3D metaphors.
  • Use kinetic typography sparingly to punctuate key technical claims before showing the visual proof.

Want a video like this for your product?

Impractical cuts one from a single prompt — the same motion craft, in about twenty minutes.