Runway launches Multi-Shot App on web for prompt-to-scene with dialogue and cinematic cuts

Runway’s new Multi-Shot App is framed as a scene builder, not just another text-to-video endpoint. In the announcement, the company says a single prompt can produce dialogue, sound effects, intentional cuts, pacing, and cinematic framing, and that the workflow supports either image-to-video or pure text-to-video generation. That matters for filmmakers and designers because the unit of generation shifts from isolated shots to a blocked sequence.

Runway also says the tool is available now on the web app, with its follow-up post linking directly to the product. The company’s examples make the pitch concrete: instead of showing one polished hero clip, the thread shows short scene fragments built from plain-language prompts, including character exchanges and timing-based beats.

The strongest pattern in Runway’s demo thread is that the prompts read more like scene briefs than camera commands. The squirrel-and-seagull example in the first demo starts from a simple comic premise, while the tension beat strips the wording down even further to "The two sit in awkward silence as the tension rises," suggesting the app is inferring coverage and pacing from narrative intent rather than requiring detailed shot lists.

The other demos broaden that range. the mice argument uses a dialogue-heavy premise with a clear comic reversal; the monster therapy scene stages multiple characters reacting inside a single setup; and the lion-on-couch clip pushes toward more photoreal character performance. Then the swamp fantasy prompt shifts gears completely, turning a long-form descriptive prompt about a humanoid toad, an old hag, and a foggy marsh into a more overtly cinematic fantasy beat.

Across those examples, the creative takeaway is less about one visual style than about structure: prompts that specify relationship, conflict, or a tiny dramatic turn seem to be what Runway wants the model to expand into cuts, sound, and scene rhythm.