Text to 3D Animation
Describe the motion.
Get the animation.
Our text-to-motion model turns natural language into 3D character animation in seconds, anything from a casual wave to a full fight combo. It lives in PINOC: write the prompt there, take the motion anywhere as FBX or GLB.
Dance, combat, sports, gestures. All from a text prompt.
Describe it like you’d brief an animator and PINOC animates it. Fast, expressive action is its home turf: moves come back big and readable, emotion included, and specific asks work too, a hand-off, a dodge into a kick.
- Crossover dribble left to right, spin past a defender, rise into a fadeaway jumper.
- A person performs a spinning roundhouse kick with a strong follow-through.
- A person performs a salsa basic step with rhythmic hip movement.
- A person plays an imaginary guitar with strumming and body rhythm.
How to turn text into 3D animation in three steps.
- 1Step 1
Write the prompt, set the take.
Type the move, set the clip length, and choose how many variations you want back. Need a specific pose? Drop an image in as the first or last frame.
- 2Step 2
Preview and pick.
Every run returns multiple takes. Play them on a character model to preview and keep the best one.
- 3Step 3
Export FBX or GLB.
Motion data ready for Blender, Unity, Unreal, or any 3D tool.
Previs, game prototypes, NPCs. Where generated motion goes to work.
Block shots and previs scenes.
Motion comes back fast enough to work like sketching: rough pass, play it on your character, reword, run it again. Approved blocking is already exportable animation.
Prototype gameplay in Unity or Unreal.
A new mechanic needs an attack, a dodge, and a hit reaction before you know if it’s fun. Prompt them in minutes, drop FBX or GLB into your engine, playtest with real motion instead of a sliding T-pose.
Animate NPCs and background characters.
Hero characters earn hand-keyed polish. The other twenty just need to behave: cheer in the crowd, argue across the bar, celebrate the goal. No reference footage, no preset? Describing it is enough.
In PINOC today. In your pipeline next.
Text to motion runs inside PINOC right now, and you can try it for free. The same model is coming to our API, so studios and tools can generate motion programmatically.
Explore the APICommon questions.
What is text to animation AI?
Text to animation AI turns a written description into 3D character animation. Type what the character does and PINOC generates the motion. No keyframes, no reference footage. Preview it on your character and export when it reads right.
How is it different from motion capture?
Motion capture reproduces a performance you can film. PINOC’s video mocap pulls motion from real footage, while text to motion generates motion from a description. Use mocap when you have the performance on camera, and text to motion when the move only exists in your head.
How is it different from Mixamo?
Mixamo gives you a fixed library of preset clips. Text to motion generates the exact move you describe, so you’re not limited to the closest match. Exports use a Mixamo-named skeleton, so both fit the same pipeline.
What formats can I export?
FBX and GLB, on a 65-bone Mixamo-named skeleton. Works in Blender, Unity, Unreal, Maya, or any other 3D tool.
How fast is it, and how many variations do I get?
A run finishes in seconds and returns up to 4 variations in a single batch. Set the clip length and how many you want before you run, then preview them on a character and keep the one that reads best.
Can I control the starting or ending pose?
Yes. Drop an image in as the first or last frame and the motion starts or ends on that pose. Handy for matching the previous shot or landing a cut exactly where you need it.
Can I use it via API?
Not yet. Text to motion is available in PINOC today, and API access is on the roadmap. If you want to build with it, contact us and we’ll keep you in the loop.
Is it free?
There’s a free trial. Sign in, try it for free, and upgrade when you need more. Send us your feedback while you’re at it, it shapes what we build next.
How do I write good prompts?
Name the move, then say how it is performed: "a spinning roundhouse kick with a strong follow-through" beats "a cool kick". Chain actions in the order they happen, keep one idea per prompt, and when a take is close, change a single word and rerun instead of rewriting the whole thing.