A field line, a swinging pendulum, a wave passing through a medium: none of these are static objects, yet most physics classes teach them as static drawings. Manim, and the manim-physics plugin built on top of it, can actually simulate them. Here's what that looks like, and how to make it without writing Python.
Electric field lines between two point charges, simulated with manim-physics and generated in Animo from a plain-language description.
Math has functions and shapes that at least hold still long enough to sketch. Physics is almost entirely about things that don't: a ball accelerating under gravity, a field pushing a charge, a wave carrying energy through a medium it never actually moves. A textbook can only show you one frozen instant of all of that.
So teachers compensate. Arrows get added to imply direction. Dotted lines imply motion that already happened. Students are asked to imagine the part that matters most, the part where the diagram just stopped drawing.
Manim was built to close this exact gap for mathematics. Physics needs it even more, because velocity, force, field strength, and phase are all defined by how something changes, never by what it looks like sitting still.
Core Manim already handles a lot of physics: vector fields, kinematics graphs, orbit paths, Bohr-model atoms, anything you can describe as shapes moving along a scripted path. manim-physics goes a step further: it's a community plugin that wires an actual 2D physics engine (pymunk) into Manim, so objects fall, collide, and settle under real simulated forces instead of a hand-keyframed guess at what gravity looks like.
A SpaceScene runs a real physics simulation: balls fall, bounce, and collide under actual gravity. Pendulum and MultiPendulum classes simulate single and coupled pendulums, including chaotic motion that would be nearly impossible to keyframe by hand.
Charge and ElectricField classes draw the field lines a point charge actually produces. MagneticField and Wire simulate the field around a bar magnet or a current-carrying conductor, the two diagrams every intro E&M course draws by hand and gets slightly wrong.
LinearWave and RadialWave animate propagation through a medium. StandingWave renders the fixed nodes and antinodes that show up on a string or in a resonating cavity, the exact pattern that's nearly unreadable as a still image.
Lens and Ray classes trace how light actually bends through a lens, showing refraction and focal points as paths instead of a single labeled diagram of where the image forms.
You don't have to choose one or the other. A single scene can mix a manim-physics rigid-body simulation with a hand-drawn graph, a MathTex equation, and a Bohr-model atom built from ordinary circles and dots. The plugin only steps in where a real physics engine is actually needed.
Most Manim research so far looks at math and computer science classrooms, not physics. But the core finding, that animation beats static graphics for teaching anything that unfolds over time, applies to physics just as well, maybe better, since forces, fields, and waves are nothing but change.
Christina Zhang's 2025 paper Manim for STEM Education: Visualizing Complex Problems Through Animation (arXiv:2510.01187) looked at Manim across STEM courses and found animations consistently beat static graphics on learning outcomes, with viewers able to follow multi-step processes in a way reading the same steps never quite managed.
Zhang, C. (2025). arXiv:2510.01187
A 2023 pilot study at Griffith University built Manim animations for convolution and Fourier series, wave physics in every meaningful sense, for a Signals and Systems course, then surveyed the electrical engineering students who watched them. Nobody in the group of 15 disagreed the animations helped, and about half said they understood the concepts better than from the lectures alone.
So, S., Schwerin, B., Rowlands, D., Espinosa, H., Tadj, T. & Busch, A. (2023). Proceedings of the 34th Australasian Association for Engineering Education Conference (AAEE 2023), Griffith University
A 2025 systematic review in the European Journal of Science and Mathematics Education (13:4) looked at visualization across secondary school STEM subjects. Students working from visual representations consistently outperformed those working from symbolic notation alone, with animation the most effective strategy of the reviewed approaches.
European Journal of Science and Mathematics Education, 2025, 13(4), 352–367
Some physics can be taught from a still frame. Most of it can't, because the thing being taught is the change itself.
A ball falling, bouncing, and colliding under simulated gravity shows energy transfer directly. A parabola on a whiteboard only shows where it ends up.
Simple harmonic motion is intuitive once you watch it swing. Coupled and chaotic pendulums are nearly impossible to reason about from equations alone.
Field lines around a point charge, or between two charges, are drawn from memory in most classrooms. Simulating them from the actual charge positions removes the guesswork.
The field around a bar magnet or a current-carrying wire, and what happens as a magnet moves through a coil, are core E&M ideas that a static right-hand-rule diagram barely conveys.
Nodes and antinodes on a vibrating string, or two waves interfering constructively and destructively, only make sense once you watch the superposition happen in real time.
Light bending through a lens, converging at a focal point, or reflecting off a curved mirror is a path, not a single labeled arrow.
Centripetal force, orbital mechanics, and why a satellite doesn't fall into what it's orbiting all land faster when you see the trajectory traced live.
Electrons orbiting a nucleus, energy-level transitions, and simplified orbital shapes are staples of intro physics and chemistry, and they're one of the easiest things to build in core Manim.
Manim and manim-physics are Python libraries. Setting either one up, let alone combining a rigid-body simulation with a hand-built field diagram in the same scene, is a real engineering task most physics teachers don't have a spare afternoon for.
Animo skips that setup. Describe the physics scenario in plain language, and it writes the Manim code, reaching for manim-physics when the concept needs a real simulation, then renders the video.
The first animation takes 20 to 40 minutes. After that, you reuse and adapt the same simulation every semester: swap the charge signs, change the pendulum length, or drop in a new wave frequency without rebuilding the scene from scratch.
Animo ships with Manim, manim-physics, LaTeX, and ffmpeg already bundled. You don't need to install Python, pymunk, or touch a terminal. Open the app and start describing.
Pick the concept your students struggle to picture: a field, a collision, a wave. Describe it to Animo and see whether a 30-second simulation changes how they respond to it.