A new tool called Motion generates full motion-design videos — scripted, animated, with voiceover and music — from a single text prompt. The team made their own launch video with it.
Motion positions itself as an agent, not a generator. You feed it a prompt plus context (product links, X threads, YouTube videos, brand guidelines), and it handles the pipeline end to end:
- Research and scripting — pulls style references and brand context from the links you provide, then writes the narrative
- Scene design and animation — storyboards individual scenes, designs them, and adds motion
- Voiceover and music — layers audio automatically
- Post-generation editing — resize, drag-and-drop, modify elements directly, or chat with the agent to iterate without regenerating entire scenes
The editing layer is the interesting part. Most AI video tools treat output as a black box — you get a rendered clip and start over if something’s off. Motion claims you can edit individual elements after generation, which would cut the re-prompting loop that eats time in tools like Runway or Sora.
The launch post on X shows the team’s own launch video, built entirely in Motion. Use cases they’re targeting: explainer videos, product launch clips, logo animations, and article-to-video conversion.
NOTE
Motion shares its name with the popular calendar/task app Motion (usemotion.com). Early community feedback on Hacker News flagged the naming overlap — worth noting if you’re searching for it.
Motion is live at motion.so and accepting users. No public pricing details yet. The product is early — 8 points and 3 comments on its Show HN — but the agent-with-editing approach is a concrete differentiator worth tracking against pure-generation tools.