Tesla Wrap Template Guide

Tesla ships 12 wrap templates and no documentation of what is in them. This is a reverse-engineered description of how the UV atlases are laid out — measured from the published PNGs, not supplied by Tesla.

This is not official documentation

Nothing here comes from Tesla. Every claim below was measured from the template PNGs that Tesla publishes, using pixel-level analysis of the alpha channel, opaque bounding boxes and horizontal-flip differences. It describes what the files currently contain. It is not a specification, it carries no guarantee, and Tesla can change the atlas layout in any repository update without telling anyone. Where we are inferring rather than measuring, we say so.

Why this page exists

Tesla's teslamotors/custom-wraps repository hands you a PNG per model and a short README about file limits. What it does not hand you is a UV map: no panel labels, no coordinates, no statement of which rectangle of the atlas ends up on the driver's door. If you want to place a stripe so it lines up across the door and the rear quarter, you are guessing and iterating.

In April 2026 someone filed issue #80, "Expose UV mapping / model metadata to enable community wrap tools", asking Tesla for exactly this. It has never been answered. In the discussion, the author of a third-party wrap editor described having reverse-engineered the mapping by hand, per model, because there was no other way to build a tool that could show you where your artwork would land.

That is the gap this page fills. It is an observational description, published openly so nobody else has to redo the measurement from scratch. If Tesla ever answers #80 with real metadata, this page should be replaced by it.

What a Tesla wrap template actually is

Each template is a UV atlas: a flat image whose coordinates map onto the 3D surface of the car. It is not a side-elevation drawing, which is why a design that looks correct as a flat image often looks nothing like you expected once applied. Three visual elements carry meaning:

All 12 templates are 8-bit RGBA PNGs. On the 11 car templates, between 53.8% and 73.6% of pixels are fully transparent — so on a typical car, well over half the canvas is dead space. The Cybertruck is the exception at 38.6%, because its islands are packed far more tightly. Either way the point holds: the paintable area is a minority of the image, and a full-canvas gradient is mostly being spent on pixels the car will never render.

Layout of the 11 car templates

The car atlases share a consistent structure. Flipping one horizontally and diffing it against itself gives 4.1% to 11.4% differing pixels across the eleven. That is a near-mirror. Inspection of where the differences fall shows they are almost entirely antialiasing along the black outline strokes, plus one genuine structural asymmetry: the charge port, which exists on one side of the car and therefore on one side of the atlas.

From that symmetry, the layout reads as:

The practical consequences follow directly. Text on the flanks has to be drawn twice. Because left and right are mirrored, lettering that reads correctly on one side of the car will read backwards on the other unless you author each side separately. A design that crosses the vertical centre wraps over the top of the car, not around its nose. And a stripe placed to run along the flank must be drawn as two mirrored strokes, not one continuous band across the canvas.

The Cybertruck template is structurally different

Cybertruck is not a variation on the car layout — it is a different atlas. The same horizontal-flip test returns a 49.9% mismatch, roughly four times the worst car template. Whatever it is doing, it is not mirroring left and right about the canvas centre the way the cars do.

Two further differences matter when you are drawing on it:

If you take one thing from this page: do not reason about a Cybertruck wrap using anything you learned from a Model 3 or Model Y template.

Structural summary of all 12 templates

Dimensions read from the PNG headers. Symmetry class from a horizontal-flip pixel difference across the whole canvas. We publish the class per template and the mismatch range across the group, because the per-template figures are only meaningful relative to each other.

Template Canvas Transparent Mirror mismatch Opaque bounding box Edge behaviour
Cybertruck 1024x768 38.6% 49.85% (0, 0, 1024, 768) Reaches all four edges
Model 3 1024x1024 59.8% 8.59% (11, 22, 1017, 1021) Inset
Model 3 Standard & Premium (2024+) 1024x1024 55.5% 10.57% (9, 21, 1020, 1014) Inset
Model 3 Performance (2024+) 1024x1024 55.7% 9.46% (8, 16, 1020, 1015) Inset
Model Y 1024x1024 53.8% 4.94% (25, 6, 1001, 1020) Inset
Model Y Standard (2025+) 1024x1024 69.7% 8.84% (135, 3, 888, 1021) Inset
Model Y Premium (2025+) 1024x1024 64.2% 6.27% (53, 6, 971, 1019) Inset
Model Y Performance (2025+) 1024x1024 62.5% 8.31% (27, 0, 996, 1020) Reaches top
Model Y L 1024x1024 58.6% 11.43% (20, 12, 1011, 1018) Inset
Model S (2021+) 1024x1024 65.0% 4.12% (72, 16, 957, 1021) Inset
Model S Plaid (2025+) 1024x1024 66.8% 4.45% (72, 4, 957, 1009) Inset
Model X (2021+) 1024x1024 73.6% 4.50% (0, 2, 854, 1024) Reaches left and bottom

All 12 are 8-bit RGBA PNGs. Measured 2026-08-05 by decoding each template to raw RGBA — transparency is the share of alpha-zero pixels, mirror mismatch the share differing between the image and its horizontal flip. 11 car templates plus the Cybertruck.

How these figures were measured

So you can check the work or repeat it against a future version of the repository:

What we still do not know

Being straight about the limits is the point of publishing this rather than dressing it up as a spec. Unknowns, all of which issue #80 would resolve:

Working with the templates

Practical takeaways, in the order they will save you the most time:

  1. Start from the exact template for your model and year. Wrong template, wrong mapping — no amount of nudging fixes it.
  2. Author at native size. See the size and dimension reference for the full table.
  3. Draw each flank separately if the design contains text or anything directional.
  4. Leave the transparent regions transparent. They cost file size against the 1 MB cap and render nothing.
  5. Expect to iterate. Even with a good mental model of the atlas, seam alignment is empirical. The mobile app's 3D preview is the fastest feedback loop available — see how to upload a Tesla wrap.

Our Tesla wrap maker loads the official template for each model straight from Tesla's repository so you are always drawing on the current file, and exports at the matching dimensions. If a wrap you built refuses to appear or maps oddly once uploaded, the troubleshooting checklist covers the failure modes in order.

Corrections welcome

If you have measured something that contradicts this — particularly per-model panel identification, or evidence about seam continuity — it belongs in issue #80, where it helps everyone building tools in this space rather than sitting in one person's notes. We will update this page as better information becomes public.

Related guides & tools

Frequently Asked Questions

Is there an official Tesla wrap UV map?
No. Tesla publishes the template PNGs but no panel map, no UV metadata and no documentation of which region of the atlas covers which part of the car. Issue #80 on the teslamotors/custom-wraps repository asked for exactly that in April 2026 and has never been answered. Everything on this page is reverse-engineered from the template images themselves.
What are the white and black areas in a Tesla wrap template?
White regions are paintable surface — your artwork replaces them. Black strokes are outlines marking the boundaries of each panel island. The surrounding area is fully transparent and is not paintable geometry at all; painting over it does nothing except make the file bigger.
Are Tesla wrap templates symmetrical?
The 11 car templates are close to mirror-symmetric left to right — flipping one horizontally and comparing gives 4.1% to 11.4% differing pixels, and most of that is antialiasing on the outline strokes plus the charge port, which exists on one side only. The Cybertruck template is not: it comes in at 49.9% mismatch, roughly four times the worst car, and is laid out differently.
Why does my design come out mirrored on one side of the car?
Because the left and right body-side islands are mirrored in the atlas. Artwork that reads correctly on one flank will read backwards on the other unless you draw each side separately. This is why lettering on a Tesla wrap needs to be authored twice, once per side.
Can I paint right up to the edge of the canvas?
Usually, but do not rely on it. On most car templates the opaque content sits inset — the Model 3 bounding box is (11, 22, 1017, 1021) in a 1024x1024 image, leaving a small transparent margin all round, and Model Y Standard (2025+) leaves 135 pixels on the left. But Model Y Performance (2025+) reaches the top edge and Model X (2021+) reaches the left and bottom, and the Cybertruck runs to all four. Check the table on this page for your specific model.
Is this documentation official?
No. It is community-derived, measured from the published template PNGs by inspecting the alpha channel, bounding boxes and horizontal-flip differences. Treat it as an observational description, not a specification. Tesla could change the atlas layout in any repository update without notice.

WhatsMyTools is not affiliated with, endorsed by, or sponsored by Tesla, Inc. Tesla, Model 3, Model S, Model X, Model Y and Cybertruck are trademarks of Tesla, Inc.