How to Make a Maths YouTube Channel Like 3Blue1Brown on a Budget
3Blue1Brown's visuals come from Manim, a Python engine its creator wrote and codes by hand. What that look actually takes, what a drawn narrated tool can and cannot do for maths, and a budget plan for a maths channel, worked through eigenvectors on a grid.
By openCanviz • October 22, 2026
8 min read
To make a maths channel like 3Blue1Brown on a budget, copy the teaching, not the pixels. 3Blue1Brown's visuals are made with Manim, an open-source Python animation engine that Grant Sanderson wrote for the channel, and each animation is coded by hand: exact grids that shear, vectors that stretch, equations that rearrange themselves. You can use Manim yourself for free, but it takes programming and a lot of hours per minute of video. The cheaper route is to put most of your effort into the script (intuition before formulas, one idea at a time) and use a drawn, narrated tool for the explanatory scenes, saving precise coded animation for the one or two shots where exact motion is the point.
What 3Blue1Brown actually is
It is worth being precise, because "3Blue1Brown style" gets used to mean several different things.
- The engine. Manim, short for Mathematical Animation Engine, is a Python library. Sanderson created it and keeps his own version; a separate Manim Community Edition was forked in 2020 and is the one most newcomers are pointed to, with better documentation. Both are open source under the MIT licence.
- The look. Dark background, a small consistent palette, smooth continuous transformations, and LaTeX-typeset equations. The motion is mathematically exact because it is computed from the maths itself.
- The teaching. This is the part that matters most. Visual intuition first, the formula afterwards as a description of what you have already seen. Questions posed before answers. Generality introduced only after a concrete case.
The look comes from code. The teaching comes from writing. On a budget, the second is the one you can match.
What Manim costs you
Manim is free in money and expensive in time. To use it, you need to be comfortable writing Python, install it alongside FFmpeg and, for typeset maths, a LaTeX distribution, and then describe every object and every movement in code. A ten-second shot of a grid shearing might be twenty lines once you know the library; your first one will take an afternoon. Experienced users report hours of work per minute of finished video, and it is easy to spend more time on an animation than on the idea it serves.
None of that is a reason not to learn it. If you enjoy programming, Manim is the most precise tool there is for maths animation, and alternatives like Motion Canvas (TypeScript) take a similar code-first approach. But it is worth knowing what you are signing up for before you decide the whole channel depends on it.
What a drawn narrated tool can and cannot do for maths
A tool like openCanviz drafts a narrated video from your script, with a drawing per beat in a style such as whiteboard. That is a different kind of tool from Manim, and being honest about the difference saves you a lot of frustration.
| Task | Coded engine like Manim | Drawn narrated tool |
| Exact continuous transformation (a grid shearing smoothly) | Yes, this is what it is for | No. It draws a before and an after, not computed motion between them |
| Precise plotted curves and coordinates | Yes, computed from the function | Approximate. Check every axis and point |
| Typeset equations | Yes, via LaTeX | Do not rely on it. Keep on-screen maths short and check every symbol |
| Intuitive diagrams (areas, arrows, labelled shapes) | Yes, but slowly | Yes, quickly |
| Narration matched to scenes | You record and edit it yourself | Drafted from your script, editable scene by scene |
| Word problems, history of maths, applications | Possible, but heavy | Well suited |
| Time per minute of video | Hours | Much less; most of the time goes into the script and the checking |
In short: a drawn tool is good at the explanation around the maths and poor at mathematically exact motion. If the insight of your video is a precise transformation, you need to compute it. If the insight is an idea that a labelled picture can carry, you do not.
Worked example: eigenvectors on a grid
Eigenvectors are a good test case because the classic visual explanation depends on motion. Here is how the episode splits between the two approaches.
Take the matrix with columns (3, 0) and (1, 2), which sends the point (1, 0) to (3, 0) and the point (0, 1) to (1, 2). Applying it to the whole plane shears and stretches the grid.
Scene 1, the question (drawn). "When you apply a matrix to the plane, almost every arrow gets knocked off its line. A few do not. Those are the eigenvectors." Drawing: a grid with several arrows from the origin.
Scene 2, before and after (drawn). The grid before, the grid after, with the two basis arrows labelled and their new positions marked. Drawing: two grids side by side, with labels (1, 0) to (3, 0) and (0, 1) to (1, 2).
Scene 3, the continuous transformation (coded, optional). The grid morphing smoothly from before to after, with most arrows rotating away from their original lines and two staying on theirs. This is the shot that justifies Manim. If you do not code it, a drawn version with three or four intermediate frames, clearly labelled as steps, still carries the idea.
Scene 4, the first eigenvector (drawn). "The arrow along the x axis stays on its line. It just gets three times longer. So (1, 0) is an eigenvector, and its eigenvalue is 3." Drawing: the x axis highlighted, an arrow of length 1 becoming length 3.
Scene 5, the second one (drawn). "There is another line that survives. The arrow (1, -1) goes to (2, -2), still on the same line, twice as long. Eigenvalue 2." Drawing: the diagonal line through (1, -1), the arrow doubling.
Scene 6, the check (drawn, short on-screen maths). "Check it: 3 times 1 plus 1 times -1 is 2. 0 times 1 plus 2 times -1 is -2. So (1, -1) goes to (2, -2)." Keep this on screen as plain arithmetic rather than full matrix notation, and check every sign.
Scene 7, why it matters (drawn). "Along those two lines, the matrix is just stretching. That is why eigenvectors make repeated multiplication easy, and why they turn up in everything from vibrating bridges to how search engines once ranked pages." Drawing: two labelled lines, then three small icons for the applications.
Six of seven scenes work as drawn scenes. One benefits from computed motion. That ratio is typical for a lot of maths explanation, and it is the basis of the budget plan.
A budget plan for a maths channel
- Script first, always. Write the intuition, the concrete example, then the general statement. Check every number and sign yourself; drafted maths narration gets signs and indices wrong.
- Draw the explanation. Use a drawn, narrated tool for the scenes that are diagrams, labelled examples, applications and recaps.
- Code only the shots that need exact motion. Learn just enough Manim to make those. Render them separately and add them to the edited video in your video editor.
- Keep on-screen maths minimal. A short expression the viewer can read in two seconds beats a full derivation. Put the derivation in the description or a linked note.
- Choose topics where intuition is the product. Why the area of a circle is pi r squared, why there are as many even numbers as whole numbers, what a derivative measures. These are the topics drawn explanation does best.
Make it
- 1
Pick a topic with one central picture
If you can sketch the key insight on a napkin, a drawn video will carry it. If the insight is a precise motion, plan one coded shot for it.
- 2
Write the script intuition first
Question, concrete example, the picture, then the formula as a description of the picture. At 150 words a minute, 900 to 1,500 words gives six to ten minutes.
- 3
Paste it into openCanviz
Choose Keep my wording so the narration is exactly your checked script, and pick the whiteboard style, where each diagram builds as the idea is explained.
- 4
Check every number, sign and label
Pause on each scene and compare it with your working. Fix anything wrong in the editor and keep on-screen expressions short.
- 5
Add any coded shot in your editor
Render the one exact animation separately, in Manim or similar, and place it in your video editor where the drawn version sits.
- 6
Publish with the full working
Put the derivation, a worked check and any sources in the description, so viewers who want the rigour can find it.
Common questions
Is it allowed to copy 3Blue1Brown's style? Manim is open source and widely used, and many channels have a similar look. What you should not copy is the specific videos, scripts or explanations. Make your own, and credit any explanation you learned from.
Do I need to learn Manim at all? No. Plenty of good maths explanation is drawn, handwritten or diagrammatic. Learn it if your topics depend on precise motion and you enjoy coding. Otherwise put the hours into the script.
Can a drawn tool write equations? Treat on-screen equations as something to check, not trust. Keep them short and plain. For heavy notation, use a typeset still made separately, or a coded shot.
How do I find an audience? 3Blue1Brown runs the Summer of Math Exposition, an annual event for people making maths explainers, and its community is a good place to see what works and get feedback. Beyond that, a clear title phrased as the question a student searches for does a lot of the work.
What about other technical subjects? The same split between drawn explanation and precise shots applies to programming and science. See how to make a coding tutorial channel without your face and how to illustrate abstract ideas.
Count your motion shots
Write the scene list for your first episode and mark each scene picture or motion. If fewer than one in four need exact motion, make the whole thing as a drawn narrated video first and add coded shots later only where they earn it. It is free to start.
Turn any concept into an animated explainer
Type an outline, get a narrated, animated whiteboard video in minutes. No design skills, no timeline scrubbing. Free to start.
Keep reading
A public research video is not a shorter paper. It leads with why anyone should care, states one finding in everyday words, says plainly what it does not mean, and runs two to three minutes. A method, a jargon table, a worked script on sound and plant growth, and the checks to make before posting.
A home-made nursery rhyme video for a toddler should be slow, short and spoken, with one calm picture per line and the rhyme said twice so a child can join in. How to choose a traditional rhyme, script it, draw it, and use it in a way that fits the screen time advice for under-fives.
A good science video for young children answers one question they actually asked, in two or three minutes, with a simplification that is never false, and ends by sending them off to try something. A worked example on why the sky is blue, with the script, scenes and a kitchen experiment.
A child's birthday video works best as a one to two minute story with the child at the centre: their name, their age, the animals or things they love. A practical plan with a worked animal birthday party script, a scene list, and the privacy checks to make before you share it.