There are four broad ways to make an illustration move: GIF or frame-by-frame animation, which swaps between a sequence of drawings; Live2D, which deforms the parts of a drawing on a flat plane; 3D conversion, which builds volumetric data from your art so you can rotate it; and AI video, which generates footage from a single image. These four aren't ranked from worst to best. You choose between them based on what you want to communicate and what you already have on hand. This article lines them up side by side and compares how each one works, what you need to prepare, roughly how long it takes, where it shines, and its honest limits.
Whether you want to make an icon sway just a little, give your character expressions and have it talk, or show it from every angle as a solid object, the right path is completely different. So let's start by sorting that out.
Note: Figmee currently offers figurine-style image generation, 3D model data downloads (GLB / 3MF) and turntable video export. Ordering physical 3D prints is Coming Soon.
"Making it move" can go in three directions
"Animating an illustration" actually splits into three directions. If you go looking for information without separating them, you'll end up reading tutorial after tutorial that has nothing to do with what you want.
- Moving it while it stays flat: the drawing itself is flat, and change is shown by lining up frames over time. Short loops like a blink, a wave of the hand, or a bouncing logo are the typical examples.
- Moving a flat drawing so it looks three-dimensional: the drawing stays flat, but its parts are deformed to make it look as if the head is turning. This is the territory Live2D covers.
- Turning it into 3D and moving it: you build a 3D model with real depth from the drawing, then actually rotate it to show every side. This is the only direction where "the back really exists."
Only the third is fundamentally different in nature. With the first two, the artist decides which angles are shown; with the third, the viewer gets to choose the angle. Put the other way around, it also means you become responsible for preparing the back.
Comparing the four methods side by side
Start by checking the overall picture in the table. The time estimates are for "someone trying it for the first time, until they produce their first result."
| Method | How it works | What you need | Rough time for a first attempt | Freedom of movement | Best for |
|---|---|---|---|---|---|
| GIF / frame animation | Swaps between a sequence of frames | Drawing software, several drawings | A few hours to a few days | As much as you draw | Icons, stickers, short loops |
| Live2D | Applies meshes to parts of the drawing and deforms them | Dedicated tool, a PSD split into parts | A few days to a few weeks | Depends on how you design the deformation | Streaming, in-game character art, expressive acting |
| 3D modeling | Builds volumetric data from the drawing and rotates it | 3D software or a conversion service | A few minutes to a few weeks | All angles, but the back must be prepared | Showcasing work, planning merch, a base for 3D work |
| AI video generation | Infers and generates video frames from a single image | A supporting service, a source image | A few minutes | You can give directions, but results vary | Mood shots, testing ideas for effects |
There's no single "right answer"; it depends on the kind of information you want to convey. A practical way to sort it: lean toward Live2D if you want to convey expressions and emotion, toward 3D if you want to convey the shape itself, and toward GIF if you want to convey tempo and timing.
Method 1: Animating with GIF or frame-by-frame animation
This is the oldest method and the one with the fewest misunderstandings. You draw several pictures and switch between them in order to create movement. Because the mechanism is simple, you have complete control over the result.
- Break down the movement (about 30 minutes) — Decide first how many frames make up one loop. Starting with around 3 to 4 frames for a blink, or 6 to 12 frames for a wave, makes it less likely to fall apart.
- Draw the key frames (about 1 to 3 hours) — Draw only the start and end points, plus the extreme poses in between, first.
- Add the in-between frames (about 1 to 3 hours) — Fill in only where the motion looks stiff. You don't need to fill everything evenly.
- Adjust the timing (about 30 minutes) — Just changing how long each frame is displayed changes the impression a lot. Keep fast movements short and hold the anticipation longer.
- Export and check (about 15 minutes) — Look at it at the actual display size and make sure there's no jump at the loop seam.
Limits: you can only show it from the angles you drew. If you want the character to face sideways, you have to draw a side-facing picture. It's also in the nature of this method that the workload grows in direct proportion to the number of frames.
Method 2: Animating with Live2D
The official manual describes Live2D Cubism Editor as "a 2D modeling tool that creates three-dimensional expression from illustrations" (source: Live2D Cubism Editor Manual, as of September 2026). The official website also presents its approach of moving your drawing directly while keeping its appeal, and notes that it is widely used in streaming, games and animation as a standard for real-time 2D expression (source: Live2D "About Cubism", as of September 2026).
The basic workflow looks like this.
- Draw the illustration split into parts (a few hours or more) — Separate layers such as front hair, back hair, eyes, mouth and body. This step largely determines how much freedom you'll have later.
- Apply meshes to the parts (a few hours) — Set up the grids that serve as the foundation for deformation.
- Design the parameters (a few hours to a few days) — Create control axes such as "face angle X" and "eye open/close," and link each part's deformation to them.
- Check the motion and refine it (a few days) — A painstaking process of eliminating the angles where things break.
Its strength is that the touch of your drawing moves just as it is. The quirks of your lines and the feel of your coloring aren't lost. In terms of how densely it can convey facial acting, too, the expressive power of this method is unique.
Its limit is that the range of movement stays within "the range in which the drawing can be deformed." Live2D data doesn't hold any shape with depth, so you can't extract a 3D model from it. The details of this structure are laid out in Why Can't You Convert a Live2D Model to 3D?. If you're also thinking about merch or a 3D version, Turning Your Live2D Character into Merch: Options to Know First is a useful reference.
Method 3: Turning it into 3D and rotating it
This direction builds volumetric data from your drawing. This one is less about "making it move" and closer to "making it exist." Once it's in 3D, you can add things later, such as rotating it, zooming in, or giving it a different pose.
There are two ways to make it. One is to model it yourself in 3D software. The leading example is Blender, a free, open-source, all-in-one 3D creation suite that covers everything from modeling to animation and rendering in a single program. It offers the most freedom, but it's also true that the learning curve up to your first model is steep. How to restart from the step where you got stuck is covered in A 3D Modeling Roadmap for People Who Gave Up on Blender.
The other is to have AI infer the 3D shape from your drawing. With Figmee, the flow is as follows.
- Upload the drawing you want to turn into 3D (about 1 minute) — A scan of hand-drawn art, a smartphone photo, or digital artwork all work.
- Generate a figurine-style image (about 2 to 3 minutes) — The AI analyzes your drawing and converts it into a single image with thickness and shading. This is the step that becomes the base for the 3D model.
- Generate the 3D model (a few minutes) — Depth is inferred from that base and 3D data is created.
- Rotate it 360° in your browser to check it (about 2 minutes) — Don't just look at the front; always check the profile and the back.
- Download the GLB / 3MF, or export a turntable video (about 1 minute) — GLB is a format used for 3D display on the web and in apps, and 3MF is a format used for 3D printing.
A note on GLB: glTF is a royalty-free specification for efficiently transmitting and loading 3D scenes and models, and .glb is the format that packs it into a single binary file. glTF 2.0 became an international standard in 2022 as ISO/IEC 12113:2022 (source: Khronos Group "glTF Overview", as of September 2026). In other words, once you have a GLB, you keep the option of bringing it into other software later.
We'll be honest about the limits too. A single drawing doesn't show the back, so the rear view is filled in by the AI's inference. It can come out differently from what you intended. Also, extremely fine lines such as strands of hair, and thin protruding parts, may get thicker or be omitted during the 3D conversion. This isn't cutting corners; it's a constraint that inevitably remains whenever you build 3D from a flat image. The details of how it works are explained in How Image-to-3D AI Works. Cases that tend not to go well are laid out in Why People Say Illustrations Can't Be Turned into 3D.
As a rough guide to pricing: with Figmee, figurine-style image generation is free for the first 5 times as a sign-up perk (valid for 3 months from registration, first registration only), and 3D model generation is free for your first model per account; after that, it costs 5 tokens ($5.99 for 5 tokens). The full picture is in How to Turn an Illustration into 3D (2026 Guide).
When you use an exported turntable video to announce your table at a convention such as Comiket, keep it separate from the menu sign you put out at the venue and attach it to posts on your own social media or online shop pages instead. This is covered together with how to make the menu sign in How to Make an Artist Alley Menu Sign.
Method 4: Generating video from a single image with AI
This method takes an image as input and generates a few seconds of footage from it. It takes overwhelmingly few steps and is powerful for making mood shots.
On the other hand, by its nature it comes with some constraints. Because the image is regenerated for each frame, lines and coloring wobble slightly. Details of the character (hair accessories, clothing patterns, the number of fingers) can change unintentionally. And because the result changes every time even with the same input, it's hard to predict how well an instruction like "I want this movement" will take effect.
So it works best not as material for reference sheets or merch, but as an eye-catching shot that makes people stop on the first image of a post. In situations where you need accuracy as a piece of work, it's safer to make one of Methods 1 to 3 your main approach.
Choosing by goal
| What you want to do | First choice | Notes |
|---|---|---|
| Make an icon or sticker move a little | GIF / frame animation | Done with only a few frames |
| Have it talk on stream or show expressions | Live2D | Keeps the touch of your drawing as it is |
| Show the shape itself from every angle | 3D modeling | You'll need to prepare the back |
| Get a base for 3D merch or sculpting | 3D modeling | Check first whether the data stays with you |
| Quickly make a single mood shot | AI video generation | Use it on the assumption that details will wobble |
Even if you choose 3D, you don't have to jump straight into learning modeling. There's a hybrid route where you get a rough version with AI conversion and then refine it in 3D software. How to go about it is covered in Is 3D Modeling Too Hard? A Shortcut to Making Your Character 3D After Giving Up and Prototype in Figmee, Refine in Blender: The Complete Guide.
The four aren't mutually exclusive
In practice, you can combine them. Here are some common flows.
- Make a 3D model, export a turntable video and use it as the first image of a post. No explanation of the movement is needed, so viewers grasp the shape in a few seconds. How to make one is covered in How to Turn a 3D Model into a Turntable Video for Social Media.
- Create facial acting in Live2D, and separately prepare a 3D model to explore a 3D version. You don't have to give up either one.
- Rough out the movement with a GIF, then move on to the final production once the direction is set.
When moving it becomes the goal in itself, you tend to get stuck comparing methods. If you work backward from "what do I want viewers to take away," the choice gets made surprisingly fast. Sorting out how to present your work for each place you share it, whether on social media, in a portfolio or as a 3D version, also gives you material for the decision.
Frequently asked questions
Q. Is 3D conversion required to animate an illustration? A. No. If you only want to show an icon swaying or a change of expression, GIF or Live2D gets you there in fewer steps. 3D conversion suits cases where you want to "show the shape itself from every angle" or "get a base for 3D work or merch." If your goal is the former, 3D is a detour.
Q. Can I convert a model made in Live2D into 3D? A. No. Live2D works by deforming meshes applied to a drawing on a flat plane, and it doesn't hold shape data with depth. If you need 3D, the flow is to build a new 3D model from the original artwork, not from the Live2D data. For details, see Why Can't You Convert a Live2D Model to 3D?.
Q. I only have one drawing. Can I still go as far as 3D? A. Yes. However, the back is filled in by the AI's inference, so the rear view may differ from what you intended. Information about the back affects the result, so preparing even a simple rough sketch makes it more stable. Drawings facing front to slightly diagonal, with a clear main subject and a simple background, tend to convert to 3D more easily.
Q. What format is safest for saving animated work? A. Keeping the source data comes first. GIFs and AI videos are hard to re-edit after export, so keep your drawing files and project files. For 3D, check beforehand whether you can download in a format that follows an international standard, such as GLB, so you can bring it into other software later.
Q. I can't decide where to start. Is there a rough order? A. The realistic approach is to finish one piece with the method that takes the fewest steps. The shorter the distance to completion, the sooner you find out what you really wanted to show. From there, if the missing information is "the shape of the back," move on to 3D; if it's "facial acting," move on to Live2D. That cuts down on detours.
Summary
The ways to animate an illustration can be sorted into four: GIF or frame animation, which moves it while it stays flat; Live2D, which deforms the drawing to show it; 3D conversion, which builds volumetric data and rotates it; and AI video, which generates footage from a single image. Each is good at different information: GIF suits tempo, Live2D expression, 3D shape, and AI video mood. 3D conversion isn't an all-purpose upgrade; it's one option that works when you "want to show even the back" or "want a 3D base." If you narrow your goal down to one before choosing a method, you'll reach your first finished piece without detours.
Related guides
- How to Turn an Illustration into 3D (2026 Guide): Bring Your Lineart to Life with AI
- Why Can't You Convert a Live2D Model to 3D? How It Works and Your Realistic Options
- How to Turn a 3D Model into a Turntable Video for Social Media
- What Is Rigging? A Gentle Guide to How 3D Models Move and When You Need It




