How it worksCompareFAQNewsCareersGet early access →
NewAnimatica awarded EU funding to advance AI for character animationAwarded EU funding→

Animate a shot in minutes,
not weeks.

Place the keyframes that matter. Animatica animates the rest.

No spam. One email when your early access opens.Watch 2-minute walkthrough
backflip_hero_v2.anim
DirectGenerateRefine
30 fps4 keyframes → 80 framesGenerated in 1.4s
Keyframe 01· Takeoff
Keyframe 02· Tuck
Keyframe 03· Peak
Keyframe 04· Landing
KeyframesAI inbetween
How it works

Direct. Generate. Refine.

A three-step loop. You set the intent, the model fills in, you stay in the loop at every frame.

01 · Direct

You set the keyframes.

Place key poses at the moments that matter. Every keyframe is a hard constraint, so your creative intent is preserved exactly. The model never overrides what you asked for.

backflip.anim
Direct
4 keyframes
02 · Generate

Same direction. Infinite takes.

From the same keyframes, generate diverse, physically plausible motion. Explore ten variations before committing to one. Each take is a different interpretation of the same intent.

backflip.anim
Generate
80 inbetweensTake 01/10
03 · Refine

Adjust anything. Regenerate the rest.

Move a pose, add a constraint, regenerate a segment. Every change propagates through the motion while preserving the keyframes you've already approved.

backflip_v2.anim
Refine
K03 editedRegenerated
Posing

Pose it in seconds. Not hours.

Keyframing starts with the pose. Block a rough shape or describe it in a sentence. Animatica lands a clean, production-ready pose on your rig.

AutoposeAutopose: rough block cleaned into a production pose
Autopose. Rough block → physically valid pose in one pass.
Text-to-poseText-to-pose: generate a pose from a short prompt
Prompt: “a person waving”
The difference

Hand-keyed tweens vs. Animatica.

Drag the divider. Same two keyframes: one is hours of hand-keying, the other is seconds.

Hand-keyed · linear tweenHand-keyed
Animatica · generated motionAnimatica
Drag to compare
Hand-keyedLinear tween between 3 poses. Stiff; no physical plausibility.
With AnimaticaFull 64-frame inbetween, physically plausible, keyframes respected
Your rig

Your rig. Not ours.

Any humanoid skeleton, any proportions, any bone naming. Realistic, stylized, anthro, armored. Animatica maps automatically. No T-pose retargeting, no rename pass, no “use our rig.”

multi-rig.anim
Same keyframes·6 rigs in sync·Pose 02 · Peak
Six different humanoid characters (stylized, fox-anthro, realistic, base mannequin, armored hulk, and mummy), all hitting the same keyframed pose on a shared stage.
Six rigs. One keyframe pass.Quadrupeds and creatures next
In your pipeline

Runs where you work.

FBX in, FBX out. Bring your rig into Proscenium, animate, then drop the result into Blender, Maya, Unreal, Unity, or any tool your pipeline already reads.

integrations.configFBX I/O
Proscenium

Proscenium

Standalone animator

BlenderMayaUnreal EngineUnity

Your DCC

Blender · Maya · Unreal · Unity

FBX in and out. Animatica lands in anything your pipeline already reads.
Before you ask

The questions we get most.

The opposite, actually. Every keyframe you set is treated as a hard constraint, not a suggestion. The model fills in around the poses you've authored. If you lock a frame, it stays locked through every regeneration.
Yes. We support standard skeletal formats (BVH, FBX, USD in and out), and the output is clean keyframe data that lands back in Maya, Blender, Unreal, or Unity. No custom rig rework, no black-box plugin.
You do. The motion generated from your keyframes is yours, full stop. Commercial rights included. The underlying model was trained on licensed and public motion capture data; we don't retain or reuse your scene data.
The model samples from a broad distribution of physically plausible motion. Each take is distinct. You can generate ten variations from one set of keyframes and pick the read that fits the shot. Style-transfer tooling is on the roadmap for alpha partners.
Mocap gives you one performance, and you're stuck reshaping it. Procedural auto-animation blends canned clips. Animatica generates fresh, continuous motion that honours the poses you draw, frame by frame. You direct; the model fills in.
For enterprise partners we offer on-prem and VPC deployments. Scene data and rigs never leave your infrastructure unless you explicitly opt into cloud inference. Full details in the access call.
Early access

A seat in the alpha, or the waitlist.

We open a handful of seats each month to studios and professional animators. Leave your email and we'll send an intro note and a private demo when a slot is yours.

✓ No spam. Only when access opens.✓ Studios and individuals both welcome.✓ NDA on request.