motionsmith-v1-docs
MotionSmith V1 Documentation
BuySetup, technical reference, features, production workflow, examples, tutorials, FAQ, and V1 support in one place. Built for Unreal Engine 5.8.
Contents
1. Quick Start Guide
MotionSmith AI provides a local motion-generation and animation-authoring workflow inside Unreal Engine 5.8. Describe a motion, generate and review the result, extend it with additional segments, refine the performance, retarget it to a supported character workflow and create a standard Unreal animation asset.
Prompt → Generate → Review → Continue → Retarget → AnimationSequence
First-Time Setup
- Install and enable MotionSmith AI Plugin for Unreal Engine 5.8 (C:\Program Files\Epic Games\UE_5.8\Engine\Plugins\Marketplace)
- Download the separate MotionSmith Runtime package.
- Extract the Runtime contents to
%LOCALAPPDATA%\MSR\. - Obtain Meta Llama 3 8B Instruct separately from the official Hugging Face repository.
- Open Editor Preferences → Plugins → MotionSmith.
- Select your local Meta Llama 3 8B Instruct model folder.
- Click Rescan and confirm that the required MotionSmith components report READY.
- Run Diagnostics / Pre-Flight and confirm all required checks pass.
Create Your First Motion
- Open MotionSmith AI Studio.
- Enter a short, focused motion prompt.
- Generate and review the motion in the viewport.
- Use Add Segment to continue the performance without rebuilding already approved motion.
- Retarget to your target character when required.
- Create the final Unreal
AnimationSequence.
Existing animations can also be imported into the MotionSmith workflow, adjusted, placed on the timeline and continued with newly generated motion.
2. System & Hardware Requirements
MotionSmith AI runs its generation workflow locally. Actual memory usage depends on the current Unreal project, viewport workload, runtime configuration and model state, so MotionSmith 1.0.0 does not publish a fixed minimum VRAM figure.
| Component | Requirement |
|---|---|
| Engine | Unreal Engine 5.8 |
| Operating System | Windows 64-bit |
| GPU | CUDA-capable NVIDIA GPU |
| Model | Meta Llama 3 8B Instruct stored locally |
| Storage | Sufficient local SSD space for the MotionSmith Runtime, Meta Llama model and project assets |
No Manual Python Environment Required
The separate MotionSmith Runtime package includes the Python environment, PyTorch/CUDA runtime, NVIDIA Kimodo-SOMA-RP-v1.1 model and the required LLM2Vec adapters. You do not need to create a Conda environment or manually install Python, PyTorch, Kimodo or LLM2Vec.
Meta Llama 3 8B Instruct is not included with MotionSmith or the MotionSmith Runtime and must be obtained separately by the user.
If GPU memory becomes constrained, reduce unnecessary GPU-heavy workloads in Unreal while generating and close unrelated GPU-intensive applications.
3. Installation & Setup
Install the MotionSmith Plugin
- Install MotionSmith AI for Unreal Engine 5.8 through Fab.
- Open the target Unreal project.
- Enable the plugin if Unreal asks you to do so.
- Restart Unreal Editor when required.
Install MotionSmith Runtime
MotionSmith Runtime is distributed separately from the Fab plugin. MotionSmith does not automatically download, install or extract the Runtime.
Download MotionSmith Runtime 1.0.0
- Download the MotionSmith Runtime ZIP.
- Open
%LOCALAPPDATA%in Windows Explorer. - Create an
MSRfolder if it does not already exist. - Extract the contents of the Runtime package into
%LOCALAPPDATA%\MSR\.
The Runtime package already includes the required Python environment, PyTorch/CUDA runtime, NVIDIA Kimodo-SOMA-RP-v1.1 model and LLM2Vec adapters.
Configure Meta Llama 3 8B Instruct
- Obtain access to the official Meta Llama 3 8B Instruct repository on Hugging Face.
- Download the model to any local folder you prefer.
- Open Editor Preferences → Plugins → MotionSmith.
- Set Meta Llama Folder to the downloaded model folder.
- Click Rescan.
Verify the Setup
The MotionSmith Setup section should report READY for:
- MotionSmith Runtime
- Kimodo-SOMA-RP-v1.1
- LLM2Vec MNTP Adapter
- LLM2Vec Supervised Adapter
- Meta Llama 3 8B Instruct
When all required components are detected, the Backend Setup Summary should report READY.
Run Diagnostics / Pre-Flight and confirm that the release gate reports READY before generating motion.
4. Model & Weights Configuration
MotionSmith Runtime already includes the NVIDIA Kimodo-SOMA-RP-v1.1 model and the required LLM2Vec adapters.
The only model that must be obtained separately by the user is Meta Llama 3 8B Instruct. MotionSmith does not redistribute the Meta Llama model weights.
Download Meta Llama 3 8B Instruct
Obtain access and download the model from the official Hugging Face repository in accordance with Meta's applicable license terms:
Meta Llama 3 8B Instruct on Hugging Face
You may store the downloaded Meta Llama model anywhere on your computer.
Select the Local Model Folder
Open Editor Preferences → Plugins → MotionSmith and set Meta Llama Folder to the downloaded model directory.
A valid model directory contains files such as:
Meta-Llama-3-8B-Instruct/ ├── config.json ├── tokenizer_config.json ├── model.safetensors.index.json ├── special_tokens_map.json ├── tokenizer.json ├── model-00001-of-00004.safetensors ├── mode2-00001-of-00004.safetensors ├── mode3-00001-of-00004.safetensors ├── mode4-00001-of-00004.safetensors └── supporting model configuration files
Click Rescan. When the model is valid, Meta Llama 3 8B Instruct will report READY.
Alternative Download Method — Hugging Face CLI
As an alternative to downloading the model files manually, you can use the official Hugging Face CLI to download the required Meta Llama 3 8B Instruct files directly into a local folder.
First, make sure your Hugging Face account has access to Meta Llama 3 8B Instruct. Then install the Hugging Face Hub CLI:
py -m pip install -U huggingface_hub
Sign in to Hugging Face:
hf auth login
Download the required model files to a folder of your choice:
hf download meta-llama/Meta-Llama-3-8B-Instruct config.json generation_config.json model-00001-of-00004.safetensors model-00002-of-00004.safetensors model-00003-of-00004.safetensors model-00004-of-00004.safetensors model.safetensors.index.json special_tokens_map.json tokenizer.json tokenizer_config.json --local-dir "C:\MotionSmith\Models\Meta-Llama-3-8B-Instruct"
When the download is complete, set Meta Llama Folder in Editor Preferences → Plugins → MotionSmith to the folder you selected, then click Rescan.
5. Backend & Pre-Flight Diagnostics
MotionSmith runs a local backend for motion generation. Setup and validation are available from Editor Preferences → Plugins → MotionSmith.
Setup Areas
- MotionSmith Setup: verifies the installed Runtime, bundled Kimodo model, LLM2Vec adapters and selected Meta Llama folder.
- Environment: validates the local Python, GPU and CUDA environment.
- Lifecycle: starts, stops or restarts the local backend when required.
- Diagnostics / Pre-Flight: performs the final environment and configuration checks before generation.
Run Diagnostics / Pre-Flight after first-time setup and after changing important runtime or model configuration. Required checks should pass before starting production generation.
If configuration has changed while the backend is already running, restart the backend so the new settings are loaded.
6. Studio Interface Overview
MotionSmith AI Studio brings generation, preview, timeline authoring, motion adjustments and retargeting into one Unreal Editor workflow.
- Authoring Viewport: preview the current character motion and spatial relationship to optional props.
- Prompt & Generate Controls: describe the next action and start generation.
- Timeline: review and manage generated or imported motion segments as a continuous performance.
- Details Panel: access generation, motion editing and retargeting controls relevant to the current workflow.
- Retargeting: transfer the working motion to a supported target character workflow without leaving MotionSmith.
7. Generation Settings
MotionSmith exposes the generation controls required for the current build directly in the Studio / Details workflow. Use the visible controls rather than modifying Kimodo or Python configuration manually.
Recommended Workflow
- Write a focused motion prompt.
- Select an appropriate Recipe when the current task benefits from one.
- Review the current generation defaults in the Details panel.
- Generate the motion.
- Judge the result in the viewport before changing multiple settings at once.
For predictable iteration, change one meaningful setting at a time. If the result does not match the intended action, simplify or clarify the prompt before making large parameter changes.
8. Recipes System
Recipes provide convenient starting configurations for different motion-generation tasks. They are workflow presets, not replacements for a clear prompt.
Using Recipes
- Select the recipe that best matches the type of motion you want to create.
- Write a concise physical-action prompt.
- Generate and review the result.
- If the motion intent is wrong, revise the prompt first.
- If necessary, test another recipe while keeping the prompt focused.
Recipe availability and exact labels are determined by the MotionSmith 1.0.0 build shown in the Studio interface.
9. Prompting Guide & Best Practices
Core Prompting Rules
- Describe motion, not story context: write what the body should physically do.
- Keep prompts focused: one clear action or beat usually produces easier-to-control results.
- Use direction and pace when useful: forward, backward, left, right, slowly, quickly, cautiously, etc.
- Use the timeline for choreography: split longer performances into multiple segments.
- English is recommended: concise English motion descriptions generally provide the most predictable text intent.
| Less Useful | Better Motion Prompt |
|---|---|
| "The soldier is terrified and searches for an escape." | walk forward cautiously while looking from side to side |
| "Do an awesome fight move." | throw a right high kick and recover to a fighting stance |
| "Walk around happily." | walk forward with a light energetic step |
Build Complex Actions as Segments
idle → walk forward → stop → turn left → step back
10. Multi-Prompt Timeline & Chaining
MotionSmith treats a longer performance as a sequence of editable motion segments. Each prompt can become its own clip on the timeline.
SEGMENT 01 idle and scan the room SEGMENT 02 walk forward cautiously SEGMENT 03 stop and turn toward the sound
Add Segment
Add Segment continues from the existing motion tail instead of regenerating the entire approved performance. MotionSmith preserves the earlier animation and uses the previous segment's transition context to generate the next motion.
This makes it practical to approve a section, continue from it, and revise the next beat without discarding earlier work.
11. Import & Continue Workflow
Existing Unreal animation can become part of the same MotionSmith working timeline rather than serving only as an external reference.
Imported Animation ↓ Transform / Adjust ↓ Add to Timeline ↓ Generate Next Segment ↓ Review Combined Motion ↓ Create AnimationSequence
This workflow makes it possible to combine authored or captured motion with AI-generated continuation and review the result as one performance.
12. In-Studio Motion Editing
MotionSmith includes practical tools for refining working motion without regenerating everything from scratch.
- Timeline Editing: work with individual motion segments and their placement in the performance.
- Transform Adjustments: reposition or reorient imported and working motion when required.
- Pose Controls: adjust a pose and hold it when the workflow requires a fixed pose decision.
- Mirror Pose: mirror a pose across the character's local sagittal plane.
- Mirror Animation: create a mirrored animation variation while preserving the intended trajectory behavior.
- Root / Path Authoring: refine character travel and root-path decisions using the controls available for the current clip.
The goal is to keep approved animation intact and make targeted corrections instead of returning to full regeneration for every change.
13. Props & Spatial Context
Props can be loaded into the MotionSmith authoring viewport as spatial references while building or reviewing character motion.
- Load the prop into the MotionSmith viewport.
- Position it relative to the character.
- Generate, import or edit the character motion.
- Review the spatial relationship in the viewport.
- Adjust the motion or prop placement as needed.
Props provide authoring context. MotionSmith's primary output remains the character animation; the presence of a prop does not imply automatic hand attachment or guaranteed physical contact solving.
14. Skeleton Retargeting
MotionSmith works with motion generated on the NVIDIA Kimodo SOMA character representation and transfers the finished motion to target Unreal characters through Unreal Engine retargeting workflows.
Included Target Workflows
- Epic / Manny-based characters
- MetaHuman
- Character Creator
- Mixamo-based characters
Typical Retarget Workflow
- Generate or import and approve the source motion.
- Open the MotionSmith retargeting controls.
- Select the appropriate target workflow.
- Select the target character / skeletal mesh.
- Run the retarget operation.
- Review the result on the target character before final output.
Characters outside the included profiles may require a separate Unreal IK Rig / IK Retargeter setup and validation.
15. Animation Output
MotionSmith creates standard Unreal Engine animation output so the finished performance can continue through normal production tools.
MotionSmith → AnimationSequence → Unreal Production Pipeline
The resulting animation can be used with standard Unreal workflows such as:
- Sequencer
- Animation editing tools
- Animation Blueprints
- Montages and state-machine workflows
- Further Unreal retargeting or cinematic production
Multiple generated segments, or imported and generated segments, can be reviewed as one continuous performance before creating the final asset.
16. Troubleshooting & FAQ
MotionSmith Runtime Is Not Detected
Confirm that the Runtime was extracted to %LOCALAPPDATA%\MSR\
and that %LOCALAPPDATA%\MSR\runtime.json exists.
Then return to Editor Preferences → Plugins → MotionSmith and click
Rescan.
Meta Llama 3 Model Is Not Detected
Make sure Meta Llama Folder points to the model root containing the configuration, tokenizer and weight files. Then click Rescan and run Diagnostics / Pre-Flight again.
Backend Is Not Ready
Open Editor Preferences → Plugins → MotionSmith and confirm that the Runtime, Kimodo model, both LLM2Vec adapters and Meta Llama all report READY. Then run Diagnostics / Pre-Flight.
Configuration Changed but the Backend Uses the Old State
Restart the MotionSmith backend from its Lifecycle controls so the updated runtime or model configuration is loaded.
Retargeted Motion Looks Incorrect
Confirm that the selected MotionSmith target workflow matches the target skeleton and that the correct skeletal mesh is selected. For custom setups, validate the Unreal IK Rig / IK Retargeter independently.
The Generated Motion Does Not Match the Prompt
Simplify the prompt and describe one physical action at a time. Split longer choreography into separate timeline segments.
The Next Segment Needs a Different Direction
Keep the approved earlier motion and revise the next segment prompt rather than rebuilding the entire performance.
GPU Memory Is Constrained
Reduce unnecessary GPU-heavy editor workloads while generating and close unrelated GPU-intensive applications. Memory use varies with the current Unreal project and runtime state.
17. Third-Party Licenses & Credits
NVIDIA Kimodo
MotionSmith AI is powered by the open-source NVIDIA Kimodo motion-generation project. MotionSmith AI is an independent product and is not affiliated with or endorsed by NVIDIA Corporation.
Meta Llama 3
Built with Meta Llama 3. MotionSmith AI does not distribute the Meta Llama 3 model weights. Users are responsible for obtaining and using the model in accordance with Meta's applicable license terms.
18. Technical Support
For bug reports, setup problems or MotionSmith AI product support:
📧 Support: motionsmithai@gmail.com
When reporting a reproducible issue, include:
- Unreal Engine version
- MotionSmith AI version
- GPU model
- The failed Diagnostics / Pre-Flight item, if any
- A relevant Unreal log excerpt
- Short steps that reproduce the problem
AI Animation Authoring, Built Around Iteration
MotionSmith AI combines prompt-driven motion generation, editable multi-beat sequencing, mocap/animation import, in-engine motion editing, and native Unreal IK retargeting inside one unified workspace.
100% Local GPU Execution
Runs directly on your NVIDIA GPU with zero cloud subscriptions, credit meters, or external API latency. Your project assets never leave your machine.
Multi-Beat Timeline
Direct performances beat by beat. Chain successive prompt segments from the exact motion tail of approved clips with seamless inertial blending.
Native Unreal Engine Export
Retarget directly to UE5 Manny/Quinn, MetaHuman, or custom skeletons. Output native
AnimationSequence assets with full Root Motion support.
Every Tool You Need to Direct 3D Performances
From single-action prompts to complex cinematic choreographies, MotionSmith provides end-to-end authoring tools without leaving Unreal Editor.
Prompt-to-Motion Synthesis
Describe kinematic actions in plain language. MotionSmith leverages Meta Llama 3 for deep spatial understanding and NVIDIA Kimodo diffusion models for fluid, lifelike motion.
Incremental Tail Chaining
Never discard good work. Add new segments that automatically inherit the momentum, foot placement, and body posture from the end of your approved motion.
Import Mocap & Continue
Bring in existing FBX, mocap data, or custom Unreal animations. Use them as the initial seed, adjust transforms, and let AI generate what happens next.
Multi-Prompt Timeline
Organize complex shots as modular prompt clips. Trim, extend, rearrange, or replace individual beats without recalculating the rest of the sequence.
Domain Motion Recipes
Select specialized generation recipes tailored for Locomotion, Idles/Gestures, Combat/Acrobatics, and Prop interactions with dedicated kinematic priors.
Interactable Props System
Place static or skeletal meshes (chairs, weapons, obstacles) into the preview viewport. Generate context-aware animations interacting directly with scene objects.
Pose Tweaking & Mirroring
Fine-tune joints, pin key poses, adjust ground trajectories, and instantly mirror full animation tracks across axes with dedicated in-engine authoring tools.
IK Retargeting & Root Motion
Transfer motion directly from the SOMA generation skeleton to MetaHumans, Manny, or custom rigs. Export true Root Motion or In-Place Animation Sequences.
AI Does Not Have to Start From Zero
Traditional AI tools force you into an all-or-nothing approach. MotionSmith bridges the gap between hand-crafted artistry, motion capture, and AI synthesis.
Import an existing animation clip, position it on the timeline, and use AI exclusively where you need to fill gaps, create transitions, or extend the performance.
Mix Authored & Generated Motion
Mocap Walk Cycle ➔ Adjust Root Transform ➔ AI Generate: "stop and draw sword" ➔ AI Generate: "take combat stance" ➔ Export to Sequencer.
Keep Your Entire Animation Pipeline Inside Unreal
Eliminate the tedious cycle of generating motions on external websites, downloading FBX files, fixing broken bone rotations, and re-importing into Unreal.
Everything happens directly inside Unreal Editor. When an animation is approved, it bakes instantly into your project's Content Browser ready for State Machines, Control Rig, and Sequencer.
Built for Production Handoff
- Zero external dependencies or web browser switching
- Direct IK Rig & IK Retargeter profile integration
- Full support for Root Motion extraction and ground alignment
- Seamless compatibility with Control Rig & Sequencer
Ready to Supercharge Your Character Animation?
Explore our step-by-step guides, technical docs, and visual prompt examples.
Direct. Chain. Refine. Retarget.
A complete breakdown of how technical animators, cinematic directors, and gameplay developers move from initial concept to engine-baked animation sequences using MotionSmith AI in Unreal Engine 5.8.
The Engine-Native Lifecycle
Zero file exports, zero web uploads. Everything stays inside Unreal Engine's memory space from prompt synthesis to Sequencer playback.
[ ENTRY LAYER ] ┌────────────────────────┐ ┌────────────────────────┐
│ Natural Language Input │ OR │ Existing FBX / Mocap │
└───────────┬────────────┘ └───────────┬────────────┘
│ │
[ SYNTHESIS LAYER ] ┌───────────▼───────────────────────────────▼────────────┐
│ Meta Llama 3 8B (Kinematic Semantics & Intent) │
│ NVIDIA Kimodo Diffusion (SOMA Skeleton Generation) │
└───────────────────────────┬────────────────────────────┘
│
[ TIMELINE LAYER ] ┌───────────────────────────▼────────────────────────────┐
│ Multi-Segment Beat Sequencing & Inertial Blending │
│ Pose Mirroring · Root Path Splines · Prop Interaction │
└───────────────────────────┬────────────────────────────┘
│
[ INTEGRATION LAYER ] ┌───────────────────────────▼────────────────────────────┐
│ Unreal Engine Native IK Retargeter (Manny/MetaHuman) │
│ Root Motion Extraction ➔ Production UAnimSequence │
└────────────────────────────────────────────────────────┘
Two Ways to Ingest and Direct Motion
Whether generating whole new scenes or filling gaps between existing mocap captures, MotionSmith handles both within the same unified timeline.
From Scratch Action Choreography
Best for rapid pre-visualization, combat moves, creature animations, and cinematic blocking where no mocap data exists.
Bridge, Transition & Extend Existing Clips
Import raw mocap or purchased animations. MotionSmith analyzes the terminal velocity and momentum, letting AI synthesize flawless transition bridges.
The 5-Stage Production Pipeline
A deep dive into how animators control each phase of the generation and baking pipeline.
01 · Direction & Kinematic Recipe Selection
STAGE 01Describe the physical motion beat with clear vectors (speed, direction, weight). Assign a dedicated Domain Recipe (e.g., Locomotion, Combat, Acrobatics, Idle/Gestures) to bias internal contact priors and joint stiffness.
02 · Multi-Beat Timeline Assembly & Inertial Stitching
STAGE 02Review the generated clip on the working SOMA skeleton. Once approved, lock the beat and hit Add Segment. MotionSmith extracts root momentum, center-of-mass trajectory, and foot ground contact from the tail frame, generating the next action seamlessly without kinematic snaps.
03 · Contextual Props & Environment Alignment
STAGE 03Load Unreal static meshes (chairs, ledges, railings, weapons) directly into the MotionSmith viewport gizmo. Synthesize spatial interactions such as climbing, sitting, drawing a holstered weapon, or vaulting over obstacles with millimeter contact precision.
04 · In-Engine Motion Polish & Trajectory Splines
STAGE 04Direct the character’s movement path in 3D space using the Root Path Spline Editor. Pin key poses across critical contact frames, mirror animations across the sagittal plane for opposite-side attacks, and engage automated IK Foot Lock to eradicate micro-slipping.
05 · Skeleton IK Retargeting & Asset Baking
STAGE 05
Map the finalized SOMA performance to your project’s target rig (UE5 Manny/Quinn, MetaHuman, CC4, or Custom
Rig). Select Root Motion extraction mode and bake out pristine, production-ready UAnimSequence
assets directly to your Content Browser.
AS_Cinematic_Encounter.uasset (60 FPS, Extracted Root Motion, MetaHuman Rig)
The Science of Inertial Blending
Traditional animation crossfading causes unnatural floatiness and lost foot plants during transitions.
MotionSmith computes second-order differential kinematics (velocity, acceleration, and center-of-mass trajectory) from the tail of Segment N to initialize the diffusion noise for Segment N+1. The character moves naturally because physics and momentum are mathematically continuous.
Why Discrete Beats Beat Long Prompts
- Granular Re-Rolls: If beat 4 fails, re-generate in 3 seconds without losing beats 1-3.
- Precision Trimming: Adjust clip in/out points on the fly with live preview.
- Per-Segment Recipe Assignment: Walk with Locomotion recipe, transition to Combat recipe for the strike.
- Zero Latency: 100% computed in local VRAM via CUDA.
Real-World Production Scenarios
See how different game and film genres utilize the MotionSmith pipeline.
Action RPG Melee Combo
An action combat designer creates a 3-hit weapon combo that roots cleanly into gameplay blendspaces.
dash forward and slash horizontally2.
spin 360 degrees into overhead heavy strike3.
backstep and reset to combat ready stance
MetaHuman Narrative Scene
A cinematic director stages a tense dialogue scene with prop-assisted sitting and nervous body language.
pace back and forth with hands on hips2.
walk towards chair and sit down heavily3.
lean forward, resting elbows on knees
Mocap Falloff Recovery
A mocap actor stepped out of the capture volume during a stunt. MotionSmith synthesizes the missing recovery.
[Imported Mocap: High fall impact]2.
AI: roll forward onto shoulder to absorb impact3.
AI: spring back onto both feet immediately
Pre-Bake Checklist for Technical Animators
Follow this 4-step checklist before baking your animation asset to ensure zero pipeline regressions.
Decide if this clip requires World Root Motion (for montages and root-driven movement) or In-Place (for blendspaces/state machines).
Enable foot contact clamping to guarantee zero sliding frames during grounded weight shifts and turns.
Ensure your target character's reference pose matches the SOMA skeleton (T-Pose alignment) in your IK Retargeter.
Bake at 60 FPS for high-fidelity cinematics, or 30 FPS for standard gameplay memory budgets.
Integrate MotionSmith Into Your Studio Pipeline
Explore visual prompt examples or jump straight into the technical documentation.
Production Examples & Prompt Blueprints
Explore real MotionSmith AI prompt chains and multi-beat character performances. Each showcase pairs the exact prompt sequence with a real Unreal Engine capture.
Dual Sword Combo with Dash Recovery
A fast-paced martial arts attack sequence designed for Action RPG combat systems. Generated as a three-beat combat sequence using the Combat & Acrobatics recipe inside MotionSmith AI.
dash forward 2 meters and slash horizontally from right to left
Beat 02 [4.0s]: jump 360 spin into heavy downward dual-blade strike
Beat 03 [3.0s]: backstep quickly into low defensive guard stance
Crouched Infiltration with Corner Peeking
A four-beat stealth performance built from focused tactical prompts: low crouched movement, a controlled halt, corner observation, and a return to cover.
crouch low and creep forward slowly with hands ready
Beat 02 [2.5s]: suddenly halt and lean against wall to the right
Beat 03 [3.0s]: carefully peek head around corner and scan ahead
Beat 04 [2.5s]: pull back into cover and signal stop with left hand
Sword Combat Sequence with Recovery Stance
A three-beat sword combat performance built inside MotionSmith AI. The sequence moves from a guarded draw into chained blade attacks and finishes with a defensive recovery stance.
draw the sword and step forward into a guarded combat stance
Beat 02: slash diagonally from upper right to lower left and follow with a quick horizontal cut
Beat 03: step back, raise the sword defensively, and settle into a ready recovery stance
Standing to Prone Firing Sequence
A two-beat tactical combat performance showing a soldier transitioning from a standing weapon-ready position into a prone firing pose, followed by a grounded weapon-handling and aiming action.
A soldier standing with a weapon in his hand lies down on the ground and takes a shot to fire.
Beat 02: A soldier lies down to fire and grabs a weapon.
Write Clear Physical Motion Prompts
Motion prompts work best when they describe observable body movement, direction, pace and posture. Use these examples as building blocks rather than trying to describe an entire scene in one prompt.
Speed & Pace
slowly creep/cautious pacebrisk walk/jog at moderate speedexplosive sprint/sudden dashgradually decelerate to a halt
Posture & Weight
crouch lowupright military postureshift weight onto left legstumble backward off-balance
Rotations & Turns
pivot 90 degrees to the leftturn 180 degrees around right shoulderstrafe diagonally while facing forwardspin 360 degrees mid-air
Combat & Hands
raise both hands defensivelythrow a sharp right straight punchlow sweeping leg kickrecover back to a guarded stance
Build Your Own Motion Sequence
Read the technical setup guide, explore the production workflow, or watch the growing Motion Examples library on YouTube.
Learn MotionSmith AI inside Unreal Engine.
Practical step-by-step tutorials for generating motion, building multi-segment performances, continuing existing animation, retargeting characters, and refining motion without leaving Unreal Engine 5.8.
Create Your First Motion and AnimationSequence
Generate a focused character motion, review it inside MotionSmith AI Studio, retarget it when needed, and create a standard Unreal Engine animation asset.
Open MotionSmith AI Studio
Make sure the MotionSmith backend is ready, then open the Studio tab inside Unreal Editor. If this is your first run, complete setup and Diagnostics / Pre-Flight first.
Write One Clear Motion Prompt
Describe the physical action you want the character to perform. Keep the first test simple and focused.
Generate, Review and Create the Asset
Generate the motion and inspect it in the viewport and timeline.
If the result is approved, retarget to the required character workflow and create the final Unreal
AnimationSequence.
Build a Multi-Segment Performance
Create longer performances one beat at a time. MotionSmith lets you approve existing motion, then continue from its tail instead of rebuilding the complete animation.
SEGMENT 01 idle and scan the room
SEGMENT 02 walk forward cautiously
SEGMENT 03 stop and turn toward the sound
Import Existing Animation and Continue with AI
Use an Unreal animation you already have as the beginning of a performance, adjust it inside MotionSmith, then generate what happens next.
Choose an Existing Animation
Start from an Unreal AnimationSequence already available in the project.
If the source came from FBX or mocap, import it into Unreal first using the normal Content Browser workflow.
Import and Adjust
Bring the animation into MotionSmith, review its position and orientation, and use the available transform controls when the motion needs to be repositioned before continuing.
Add the Next Generated Segment
Place the imported motion into the working timeline, write the next physical action, then use Add Segment to continue from the imported animation tail.
Retarget Motion to Your Unreal Character
Transfer the approved MotionSmith performance to one of the included target workflows: Epic / Manny-based characters, MetaHuman, Character Creator, or Mixamo.
| Target | MotionSmith 1.0.0 |
|---|---|
| Epic / Manny-based | Included |
| MetaHuman | Included |
| Character Creator | Included |
| Mixamo-based | Included |
Refine Motion Without Regenerating Everything
Use MotionSmith's editing tools to make targeted changes after generation: transform adjustments, pose decisions, mirroring, root/path authoring and prop-based spatial review.
Transform Adjustments
Reposition or reorient imported and working motion when the character needs a better starting placement or facing direction.
Pose Hold
Adjust a pose and use the hold workflow when a specific pose needs to remain fixed across the chosen interval.
Mirror Pose / Mirror Animation
Mirror a pose or complete animation to create left/right variations while keeping the edit inside the MotionSmith workflow.
Root & Path Authoring
Use the available root/path controls to refine where the character travels instead of regenerating an otherwise useful motion.
Props as Spatial References
Load and position a prop in the authoring viewport to judge the character's movement against an object or scene reference.
Preserve Approved Motion
Prefer targeted edits or a revised next segment over regenerating the entire performance when only one decision needs to change.
Build the motion. Review it. Keep what works.
Use the Examples page for prompt ideas, the Documentation page for complete technical reference, and the MotionSmith AI YouTube channel for video tutorials and workflow demonstrations as they are published.
Straight Answers, Without the Marketing Fog.
Clear, technical answers about AI generation, hardware limits, mocap import, IK retargeting, model licensing, and Unreal Engine 5.8 integration.
1. AI Generation & Timeline Directing
How text prompts turn into continuous 3D performances.
What exactly is MotionSmith AI?
MotionSmith AI is a local character animation studio plugin built natively for Unreal Engine 5.8. It unifies natural language motion synthesis, multi-beat timeline authoring, FBX/mocap import, joint tweaking, and Unreal IK retargeting inside one focused editor tab.
Does MotionSmith generate animations purely from text prompts?
Yes. Describe physical actions in plain English (e.g.,
sprint forward 3 meters, slide into cover). The embedded diffusion backend generates realistic 3D
skeletal motion directly matching your kinematic direction.
Can I build a long cinematic performance from multiple short prompts?
Yes. The Multi-Prompt Timeline is specifically architected for this. Rather than gambling a 15-second sequence on one giant paragraph, you direct beat by beat (e.g., Beat 1: Idle ➔ Beat 2: Walk ➔ Beat 3: Jump) and sequence them continuously.
Does adding a new segment regenerate or overwrite my earlier approved work?
No. MotionSmith uses incremental inertial tail chaining. Once you approve a clip, its keyframes are locked. New segments sample the terminal momentum, velocity, and foot contact of the previous clip without recalculating earlier work.
What are "Domain Recipes" and when should I use them?
Recipes adjust internal kinematic priors and contact stiffness. Use Locomotion for clean walking/turns without foot sliding, Combat & Acrobatics for high-flying kicks and rolls, Idle & Gestures for expressive upper-body acting, and Props Interaction for object-aware poses.
2. Mocap Import, Retargeting & Engine Output
How generated motion integrates with your game characters and assets.
Can I import existing FBX / Mocap animations into the timeline?
Yes. You can import any existing Unreal AnimSequence or FBX mocap asset. It places directly onto
the timeline and can be adjusted with Transform tools or extended with new AI-generated motion.
Which character skeletons can I retarget to?
MotionSmith includes native Unreal Engine IK Retargeter presets for:
- UE5 Mannequins (Manny and Quinn)
- MetaHumans (Male and Female body types)
- Character Creator 4 (CC4 / AccuRIG)
- Mixamo Biped Skeletons
- Custom Proprietary Skeletons via your own IK Retargeter assets
Does MotionSmith support Root Motion extraction?
Yes. When baking to an AnimationSequence, you can choose between:
- Root Motion (World Space): Injects character ground displacement into the Root bone for gameplay montages and locomotion.
- In-Place (Zeroed Root): Keeps the character anchored at origin for blendspace or procedural locomotion systems.
What assets does MotionSmith produce?
It outputs standard, native Unreal Engine UAnimSequence assets directly into your Content
Browser. They are immediately compatible with Sequencer, State Machines, Animation Montages, and Control Rig.
Can I edit key poses and mirror animations inside the plugin?
Yes. The Studio includes built-in authoring tools: Joint Rotation Tweakers, Pose Pinning, 1-Click Sagittal Plane Mirroring, and a 3D Root Path Spline Editor to redirect character walking trajectories.
Does MotionSmith replace traditional keyframe animation?
No. It serves as a rapid motion accelerator. It eliminates tedious blocking from scratch while leaving full creative control to animators for final polish in Unreal Engine.
3. GPU Requirements, Models & Local Inference
Technical requirements, local storage, and execution architecture.
Does motion generation run 100% locally on my machine?
Yes. All AI inference (text encoding and diffusion generation) executes locally on your GPU. No data, prompts, or assets are ever uploaded to cloud servers.
Do I need to install Python, PyTorch, or set up a Conda environment?
No. The separate MotionSmith Runtime includes the required Python and PyTorch/CUDA environment, so no manual Python or Conda setup is required.
What are the hardware and GPU requirements?
You need an NVIDIA RTX GPU with CUDA support:
- Minimum: 8 GB Dedicated VRAM (e.g., RTX 3060, RTX 2070, RTX 4060)
- Recommended: 16 GB Dedicated VRAM (e.g., RTX 3080, RTX 4070 Ti, RTX 4090)
- System RAM: 32 GB+ recommended
- OS: Windows 10/11 64-bit
Why do I need to download Meta Llama 3 8B Instruct separately?
Meta Llama 3 8B Instruct serves as the semantic text encoder for kinematic prompts. Due to Meta's gated license terms, third-party software cannot redistribute model weights. You simply download the weights once from Hugging Face and point MotionSmith to your local folder.
What is NVIDIA Kimodo's role in MotionSmith?
MotionSmith builds upon the open-source NVIDIA Kimodo diffusion framework, wrapping it into a production-ready Unreal Engine 5.8 workflow with timeline authoring, IK retargeting, and in-engine authoring tools.
How do I test if my setup is ready before generating?
Open Editor Preferences ➔ Plugins ➔ MotionSmith and click Run Pre-Flight Diagnostics. It automatically tests your CUDA drivers, embedded runtime, and Llama 3 model path integrity.
4. Commercial Rights, Marketplace & Support
Licensing terms, Fab availability, and customer care.
Can I use generated animations in commercial games and films?
Yes. All UAnimSequence assets generated using MotionSmith in your Unreal Engine
projects are 100% royalty-free for commercial video games, virtual production, cinematics, and interactive
experiences.
Where can I buy MotionSmith AI?
Purchase MotionSmith V1 through the Epic Games Fab Marketplace. Use the Buy button to open the V1 store section and continue to the official Fab listing.
Which Unreal Engine versions are supported?
MotionSmith AI v1.0 is compiled and fully verified for Unreal Engine 5.8. Updates for future engine releases will be provided via Fab.
How do I get technical support or report bugs?
For setup assistance, pipeline questions, or bug reports, email our engineering team directly at
motionsmithai@gmail.com. Please attach your Unreal Engine log
from Saved/Logs/MotionSmith.log for faster diagnostic turnaround.
Explore Our Detailed Guides
Check the step-by-step workflow guide or view real-world prompt choreography examples.
MotionSmith AI Support
Need help with installation, Runtime setup, motion generation, retargeting, Diagnostics, or another MotionSmith AI workflow? Start with the documentation below or contact us directly for technical support.
Find the answer quickly.
Installation, setup, generation, retargeting, and common troubleshooting topics are covered in the MotionSmith AI documentation and guides.
Technical Documentation
Complete reference for installation, Runtime configuration, models, Diagnostics, generation, editing, retargeting, and animation output.
FAQ & Common Issues
Quick answers to common questions about setup, hardware, local inference, Unreal Engine integration, licensing, and support.
Step-by-Step Guides
Practical MotionSmith AI workflows covering setup, motion generation, multi-segment animation, retargeting, and motion refinement.
Please send your logs and screenshots.
If you are experiencing a technical problem, please attach the relevant Unreal Engine log file and screenshots showing the issue when contacting support. This is especially important for installation, Runtime, model detection, Diagnostics / Pre-Flight, backend, generation, and retargeting problems.
Your Unreal project logs can normally be found inside the project's
Saved/Logs folder. If Diagnostics / Pre-Flight reports a failure,
please also include a screenshot showing the failed check.
What to include in your report.
Providing the information below makes it much easier to reproduce and diagnose technical problems.
Your Environment
- Unreal Engine version
- MotionSmith AI version
- GPU model
- Failed Diagnostics / Pre-Flight item, if applicable
Logs & Screenshots
- Relevant Unreal Engine log file from
Saved/Logs - Screenshots showing the problem or error message
- Screenshot of any failed Diagnostics / Pre-Flight check
- Short steps explaining how to reproduce the problem
What we can help with.
Technical support covers MotionSmith AI installation and product workflows.
Installation & Runtime
Plugin installation, MotionSmith Runtime setup, model detection, configuration, backend startup, and Pre-Flight Diagnostics.
Generation & Workflow
Motion generation, imported animation, continuation, multi-segment workflows, motion editing, and animation asset creation.
Retargeting & Output
MotionSmith retargeting workflows, target character issues, generated animation output, and reproducible MotionSmith AI bugs.
Still need help?
Send your issue to motionsmithai@gmail.com. For technical problems, include your Unreal Engine log and screenshots so we can investigate the issue properly.
Comments
Post a Comment