Direct the world you want to live in.
San Francisco
Joined May 2024
- Tweets179
- Following110
- Followers1.3K
- Likes334
Pinned Tweet
Reintroducing Intangible.
The difference between being misunderstood and getting greenlit.
intangible retweeted
I built this harbor once in 3D, then shot everything from it.
I blocked the set, the two people and six cameras in Intangible, framed each shot there and rendered a still from every camera.
Then I swapped in my own cast with Nano Banana Pro and animated the frames with Seedance 2.5.
The last three shots use the same camera at three points in time, so the place never changes.
3D set and cameras → @intangibleai
Cast, stills and motion → @runwayml
Unofficial spec test. Not affiliated with Stanley.
🤖 Made with AI
A simpler way to plan and ACTUALLY control your action sequences with AI:
1. Set where each vehicle goes
2. Adjust when and how it moves
3. Preview the action as you refine it
Use Motion Paths in Intangible to guide the action in your AI video.
Create a car chase with control over the action and camera:
1. Build the scene in Intangible
2. Animate the cars and helicopters
3. Set your camera angles and movement
4. Render with Seedance 2.5
Made by Anjoe Jacob, Applied AI Engineer at Intangible.
For our latest short film, Solara City, we used Intangible to:
1. Build the sets
2. Design the characters
3. Create the visual references for video generation
Wan 3.0 for video, Suno 6 for music, and Filmora for the final edit.
Decide exactly how your characters escape danger before you hit generate:
1. Draw the character’s path
2. Set how the camera follows
3. Turn the 3D animation into AI video
Build the action with Motion Paths in Intangible.
Get the shot you want without setting up a physical shoot:
1. Create the setting from a reference image
2. Arrange the people and props in 3D
3. Choose your camera angles
Then use AI to turn your spilled ramen into 4K visuals.
Turn your character design into an animated scene with control over its look and movement.
Apply a character sheet, arrange the scene and animate the action in 3D, then visualize it with AI.
Get the camera angles you want from a single reference image, without building the scene by hand.
Intangible gives Astra the 3D models and tools to assemble environments and set up cameras directly in the browser.
Try Astra with Intangible today through MCP using your own tokens.
Arrange the stage, seating and production setup in 3D
Then use AI to visualize how the space could look and feel when the lights come on and the crowd arrives
Build the scene you have in mind without modeling everything from scratch:
1. Start with a ready-made space
2. Arrange the furniture
3. Set the pose
4. Add visual references to individual objects
Your layout and look built with 3D + AI.
Build a 3D world. Direct it through conversation.
> Describe your scene
> Position the camera
> Explore 8 lighting moods
Turn your creative direction into a 3D scene with @ChatGPT + Intangible MCP
Direct the action before you hit generate:
1. Draw your movement path
2. Refine the action
3. Set the timing
Decide where & how things move with AI + 3D.
This camera move started with a line I drew 🤯
Motion Paths lets you choreograph how cameras and objects move through 3D space.
🚨 ICYMI: WE VISUALIZED @SpaceX STARSHIP CHOPSTICK CATCH 🚨
3D for the
- flight path
- camera moves
- timing
and @imagine for the final render
@HeironymusFox breaks down how he made it using @grok in Intangible
This is what it looks like when you direct AI, instead of AI directing you.
Our Creative Design Lead, @dana_shaviv, dives deep into the power of Gaussian Splats in Intangible.
You can turn a place you’ve filmed IRL into a 3D set and direct entire sequences in it.
intangible retweeted
Stop describing the move. Start drawing it.
Motion Paths in @intangibleai lets you draw the route your camera and objects follow with precision.
Draw the exact route on real geometry, set the line, and watch your scene follow your choreography.
Here’s how ⤵️
🤝 Paid partnership
When your scene is real 3D, the camera works like a camera.
Aim it, lock it on a subject, have it move on the exact path you drew out.
The shot that renders is the shot you called.
My cofounder @cmigos (ex-Apple & Unity) shows how to animate cameras in Intangible's semantic scene architecture.
Watch him stage the action, track a subject, and move the camera in 3D like he's directing on set.
Then diffusion models render the precise shot he wants.
To the creatives and founders who've felt misunderstood trying to convey a big idea, this one's for you.
