1080 × 1920
Safe zones for TikTok, Reels and Shorts: where your text actually survives
Compare official TikTok, Reels and Shorts safe-area templates in pixels, find the shared box for text, and understand why published measurements differ.
A vertical frame can be perfectly exported and still lose its price, caption or logo behind the app interface. This guide turns the platforms' current published overlays into working coordinates for an editor using a 1080 × 1920 canvas, then finds a central rectangle that survives TikTok, Reels and Shorts. It is for anyone positioning the words that make an ad understandable, not for guessing how a particular phone will crop tomorrow's placement.
What the interface takes away: TikTok#
A safe zone is the part of your video where a platform recommends keeping important information clear of its controls. The account name and caption can occupy the lower frame; action buttons occupy a side rail; the CTA and other placement-specific cards can cover more. It is not a promise that every pixel outside the zone is invisible, nor that everything inside it is guaranteed legible. The safe region is a working margin for key elements, not a crop tool.
TikTok does not publish a single numeric safe-zone table in its in-feed help text. It supplies downloadable templates and says the visible area depends on orientation, caption length and additional formats. TikTok also warns that the preview can differ slightly from the live ad.[1][2] The measurements here come from TikTok's In-Feed Standard downloadable template, not from a universal line in the help center. The file's printed measurements use a 720 × 1280 reference. Scaling each edge by 1.5 gives the following 1080 × 1920 working geometry; the alpha-channel measurement agrees with the labels.[3]
660 × 1020 px
Without the action rail, the central transparent region runs from x = 120 to 960 and y = 240 to 1260. In the rail's lower band, the right edge steps inward to x = 780, leaving a 660 × 1020 px rectangle clear of both the margins and rail. That narrower rectangle is the reliable one for a price or caption occupying the rail's height. The top strip, above y = 840, can use more width for a carefully placed headline, provided you check the real ad preview. These pixel coordinates are our derived measurement of a published template, not a TikTok-published numeric spec.[3]
There is a second reason not to hard-code the box into every cut. TikTok's older in-feed with Anchor template contains different variants for caption length. Its measured bottom reserve, doubled from its 540 × 960 reference, grows from 812 px for one line to 1014 px for four lines at 1080 × 1920. It is an older, different placement template, not a correction to the current standard one.[4] Longer caption copy or an add-on can push critical text higher than the standard example suggests.[1][2]
Meta Reels: percentages before pixels#
Meta's Instagram Reels ad guide says to leave at least 14% at the top, 35% at the bottom and 6% at each side free of text, logos and other important creative elements. The Facebook Reels guide uses approximately the same percentages.[5][6] Unlike the TikTok measurements, those numbers are Meta's published percentages. Converting them to our working canvas gives 268.8 px at the top, 672 px at the bottom and 64.8 px at either side. Round the fractional margins up to whole pixels when laying out: top 269, bottom 672, left and right 65. The corresponding clear box is approximately x = 65–1015 and y = 269–1248, or 950 × 979 px on that rounded grid. This conversion is arithmetic, not a separate Meta pixel specification.[5]
950 × 979 px
The bottom is doing the most work: account information, descriptions and the ad UI compete with burned-in captions there. Meta says its 9:16 guidance also applies to Stories, Reels and tall Instagram Feed video ads, so do not assume that moving the same export to Feed restores the space below your subtitles.[7] Meta also says that a Reels ad containing a disclaimer should leave the bottom 40% free of other key elements, rather than relying on the usual 35% reserve.[7] That is a different creative situation: revise the layout, do not shrink the disclaimer until it fits.
The percentages describe a planned composition, not the exact viewport on every device. Meta may zoom a 9:16 asset on a taller screen and crop material outside the safe area, or display it with a background.[7] Its Ads Manager provides a Safe Zone guardrail overlay. Use it to inspect the actual Reels placement, especially when an ad also runs in a mixed placement set; avoid inferring coverage from a desktop edit of the source file alone.[7]
Shorts: the right edge is expensive#
Google publishes a universal safe-area overlay for video ads and a downloadable vertical template. Its vertical overlay is 1080 × 1920. Our measurement of its transparent safe region is x = 48–888 and y = 288–1248, which matches the dimensions printed in the overlay: 48 px left, 192 px right, 288 px top and 672 px bottom. That leaves 840 × 960 px for important elements.[8][9] The pixel values are derived from Google's published image, not a claim that the Shorts interface always has precisely those controls.
840 × 960 px
The asymmetry matters. YouTube's Shorts FAQ says action buttons display over Shorts ads as they do over organic Shorts; an advertiser can turn off comments, but the other buttons still appear.[10] Google's Shorts asset page also describes a mobile CTA overlay card.[11] Keep a price, key subtitle and product label away from the right-side UI rather than centering them inside the entire 1080-pixel file and hoping the buttons avoid them. The subject's face may extend beyond the measured central rectangle; the eyes, result and any text needed to understand the offer should not.
Google's overlay is described as universal video-ad guidance, so it is a conservative composition reference for Shorts, not a frame-by-frame map of every campaign's CTA card. A horizontal asset served into Shorts can acquire blurred top and bottom areas; a proper vertical export makes the layout easier to control.[11] Check the preview for the exact campaign type. The button's timing and additional headline treatment can differ, while the published overlay remains the safer baseline.
The rectangle that survives all three#
To make a reusable master, intersect the published or measured zones rather than averaging their margins. At 1080 × 1920, choose the largest top reserve (288 from Google), the largest bottom reserve (672 from Meta or Google) and the largest left reserve (120 from TikTok). The right reserve depends on where your rectangle runs: Google's usual 192 px is narrower than TikTok's 300 px action-rail reserve in the overlapping band.[3][5][9]
660 × 960 px
The outcome is x = 120, y = 288, width = 660, height = 960: x extends to 780 and y to 1248. The TikTok rail is the reason the shared box is 660 px wide. Ignoring that rail would yield a 768 px-wide margin intersection (1080 minus 120 left and Google's 192 right), but part of it crosses TikTok's controls. These are our derived intersections of named source templates, not a platform-endorsed “one safe zone for everything.” They also cannot account for unusually long captions or a Meta disclaimer.[1][7]
This box is a useful default for words, not a mandate to fit the whole film into a narrow column. Let background, hands and product motion extend across the frame. Reserve the shared rectangle for the information whose loss changes the meaning of the ad. A face can sit a little left of center if its expression remains readable; a tiny price cannot depend on the viewer finding it beneath a social rail. If you edit platform-specific versions, recover that unused width deliberately and check each placement separately.
There is a useful distinction between safe at every point in the shot and safe only at one moment. TikTok's upper region has more horizontal room before the right rail begins, but a moving caption that falls into the rail band partway through the video still needs to clear the narrower boundary.[3] If the shot scrolls text downward or tracks a price beside a moving object, inspect its entire motion path, not only its starting position. For a shared master, keep the meaning-bearing part of that path inside the common rectangle; a platform-specific version can reclaim space after its own overlay check.
The arithmetic assumes the export really is a 1080 × 1920 portrait canvas. If you author at another 9:16 resolution, scale the coordinates by the width and height ratios, then round the protected margins inward so you do not accidentally gain unsafe pixels. For example, this article's coordinates are derived for the named canvas, not an instruction to put x = 120 on an unrelated file.[3][5][9] On a screen taller than the source ratio, Meta says it may crop or letterbox the asset; that is why a correctly scaled design still deserves a placement preview.[7]
Why published numbers disagree#
Many articles supply neat TikTok pixel margins without identifying the template, reference size, date or placement. A commonly repeated 130 / 484 / 44 / 140 top/bottom/left/right set appears in secondary guides, but the current official file measured for this article does not match it.[12][3] That secondary set is not a TikTok rule. Equally, a value that happens to work for a single caption configuration may fail with an anchor or extra UI.
| Number set at 1080 × 1920 | What it actually describes | Use it for |
|---|---|---|
| 240 / 660 / 120 / 120, plus 300 right beside rail | Derived from TikTok's current In-Feed Standard downloadable template, printed at a smaller reference size.[3] | A named standard in-feed starting overlay, checked in preview. |
| 130 / 484 / 44 / 140 | A secondary web convention; no matching TikTok file was found in the source review.[12][1] | A historical comparison, not your approval margin. |
| 269 / 672 / 65 / 65 | Derived, rounding Meta's published Reels percentages upward.[5] | Reels composition without the disclaimer variant. |
| 288 / 672 / 48 / 192 | Derived measurement of Google's published vertical overlay.[9] | A conservative vertical Google ad composition. |
Meta publishes percentages, TikTok publishes placement templates and Google publishes an image overlay. Treating all three as interchangeable “official pixel specs” erases how the measurements were obtained. The research also found a much older TikTok anchor template whose caption variants reserve substantially more of the bottom; a later standard template did not make every old placement disappear.[4][3] Cite the file or percentage, record when you downloaded it, and keep the overlay beside your export settings.
The conflict affects design decisions. A lower-third caption that survives one screenshot can cover a price in another. Rather than searching for a single number that promises permanence, isolate the critical words in the shared working box, then preview each placement with the actual caption and CTA. If the interface has moved, change your overlay and export; the layout is evidence of what you checked, not a trophy for following a blog's precise-looking table.
Place each element according to its job#
Words are not equally important. A decorative backdrop can be clipped. The price, qualification and action label cannot. Use the shared box for the elements that answer “what is this?” and “what must I know before I tap?” The short-form ad anatomy guide shows how those messages sit in the hook, proof and ask; placement should reinforce that sequence rather than turn the middle of the video into a stationary poster.
| Element | Working placement | Why it matters |
|---|---|---|
| Opening proposition | Central shared box, clear of the subject's face | It must be legible before the viewer hears the line or reads the platform caption. |
| Price or offer term | In the shared box next to the relevant product shot | A covered qualifier can change the meaning of the offer. |
| Spoken captions | Short blocks within the shared region, adjusted to the speaker and action | A full-width lower third can collide with the account description or CTA. |
| Logo | Within the shared region if recognition is essential; otherwise keep it subordinate | A corner lockup may be outside one platform's reliable area. |
| Face and product | Extend into the outer image, but preserve the eyes and decisive action | The frame should still feel full-screen; only crucial details need strict protection. |
| Legal or material disclosure | Prominent, readable and placement-checked, not simply miniaturized into a safe box | Meta's Reels disclaimer guidance reserves more bottom space; the disclosure must still be noticeable.[7][13] |
On-screen captions and platform-generated captions are different layers. If you burn words into the export, their position cannot follow a changing interface. If the platform supplies captions, inspect how they combine with the native description and CTA. The vertical caption guide addresses readable timing and contrast, which safe-zone coordinates alone cannot solve. A perfectly placed subtitle can still be too fast or too faint to read.
When the image and text want the same central space, decide which one carries the claim. If the important proof is a hand turning a control, move the caption to a clear region within the safe box during that shot; let it return only when the movement is over. If the claim depends on seeing a small indicator, use a closer shot that keeps the indicator inside the region rather than placing a large label over it. You are composing evidence and explanation together. A safe-zone overlay protects both, but it cannot decide which deserves the viewer's eye at a particular beat.
Worked example: a fictional caption and price#
Suppose a fictional kettle ad contains the three-line caption “Steady stream / over the grounds / into the cup” and a fictional $39 price. On a 1080 × 1920 master, place a compact caption block inside x = 160–720, y = 910–1080, and a separate price tag inside x = 160–420, y = 430–520. Both invented layouts lie within the derived shared box x = 120–780, y = 288–1248; the actual text dimensions depend on font and device rendering.[3][5][9]
Those coordinates are illustrative placements, not platform requirements. The caption has room to break between short phrases rather than splitting a word or obscuring the kettle's stream. The price appears when the real item is in view, rather than sitting on top of a face. If the pour passes through the caption area, shift the caption within the shared rectangle for that shot; do not force the action to happen behind opaque text. Readable line length, time on screen and contrast still need a separate check.
If the Meta version needs a disclaimer, the usual intersection is no longer enough to plan the composition. Meta's bottom-40%-free guidance changes the available lower area.[7] Position the claim and its qualification together in a newly checked layout. If a TikTok anchor or longer caption occupies more lower space, load that template and move the caption upward. Do not treat our x/y example as a guarantee against either change.
Make one still from the opening, one from the product demonstration and one from the final ask. Lay the relevant overlay over all three. In the opening, check whether the caption survives without hiding the product. In the middle, check that the viewer can actually see the pour rather than merely read a promise about it. In the closing frame, confirm the price and its terms remain together when the platform CTA arrives. These are fictional storyboard frames for the example, but the review method transfers to any product. The point is to inspect the frames where information changes, not just the prettiest frame.
Check the export, catch the common mistakes#
Start with the downloaded overlay for the actual placement and scale it to the export canvas. In the editor, check the output dimensions, overlay the transparent guide, place critical words inside its clear region and export a draft. Azelify's preview has generic safe-area guides; they are composition aids, not exact TikTok, Meta or YouTube templates. For exact placement, compare against each platform's current file or guardrail and then use its own preview.[1][7][8]
Review on a phone with the real interface visible. Read the opening caption without sound, pause on every price or qualifier, and let the CTA appear at its actual time. Check at least the placements you plan to buy rather than approving a pristine player with no description, rail or CTA. If a platform offers multiple caption lengths or creative add-ons, preview the chosen combination. An overlay from a different ad type cannot represent controls it never included.
Common mistakes#
- Calling a measured template a permanent platform spec. Record the file name, reference size and measurement method; recheck when the platform changes its download.[3][1]
- Averaging margins. Intersection takes the most restrictive relevant edge and checks whether a rail crosses the box. Averaging can place text inside an overlay.
- Protecting the entire image equally. Keep meaning-bearing words and decisive visual details safe, while allowing supporting footage to fill the screen.
- Using the regular Reels bottom margin for a disclaimer. Meta's disclaimer guidance asks for a larger bottom reserve.[7]
- Trusting the edit canvas more than the placement preview. Captions, CTA cards, screen shape and account chrome can change the visible result.[2][11]
- Making safe text unreadably small. Geometry prevents obstruction; caption pacing and contrast determine whether anyone can read the words.[14]
Checklist#
- Confirm the exact paid placement, canvas ratio and export size before positioning words.
- Download the current TikTok or Google template, or open Meta's Reels guardrail, for that placement.
- Put essential claims, captions, price and offer terms inside the relevant clear area.
- For one shared vertical master, start with the derived x 120–780, y 288–1248 box, then review its limits.
- Rework the layout for a long TikTok caption, add-on, anchor or Meta disclaimer.
- Let decorative footage fill the frame without burying eyes, product action or qualifiers.
- Watch the exported version on a phone, with the native description, buttons and CTA visible.
- Verify the meaning still reads without sound and that every caption is legible against its footage.
Keep reading
Captions people can actually read: a practical guide for vertical video ads
Set caption timing, line length, placement and contrast for vertical ads using broadcast guidance and accessibility ratios, then check it on a phone.
The anatomy of a short-form video ad, second by second
Storyboard a short-form video ad beat by beat: build a hook, show the problem, demonstrate proof and make one clear ask, with timing examples.