Instagram and TikTok image sizes for fashion, and the safe zones nobody tells you about
By The AURA Studio team. Published .
For fashion on Instagram, use 1080 by 1350 pixels (4:5 portrait) for feed posts, because it occupies the most vertical space of any allowed feed shape, and 1080 by 1920 (9:16) for stories and reels. In a 9:16 story or reel, keep faces, garment detail, and any text roughly within the central area, clear of about 250 pixels at the top and 420 at the bottom, where the interface sits over your image.
The three shapes that matter
| Placement | Ratio | Pixels | Why |
|---|---|---|---|
| Feed post | 4:5 | 1080 x 1350 | The tallest shape a feed post allows, so it takes the most screen as someone scrolls past. |
| Feed post, grid-first | 1:1 | 1080 x 1080 | Crops predictably in the profile grid. Choose it when the grid layout matters more than scroll presence. |
| Story and reel | 9:16 | 1080 x 1920 | Full screen. The same shape works for TikTok. |
For a garment, 4:5 is nearly always the right feed choice. A full-length look in a square frame either loses the shoes or shrinks the person to the point where fabric detail disappears, and fabric detail is what you are selling.
Where the interface covers your image
A 9:16 story is 1080 by 1920, but you never see all of it. The profile row and controls sit over the top, and the caption, the action buttons, and the audio strip sit over the bottom. Anything you place there is either covered or, worse, half covered.
- Top: keep roughly the first 250 pixels clear of faces and text.
- Bottom: keep roughly the last 420 pixels clear. This is the larger and more frequently violated zone.
- Right edge: the action buttons run up the right side in reels and TikTok, so keep the garment off that edge.
- That leaves a central band of about 1250 pixels of height that is reliably visible. Compose for that band, not for the canvas.
These are working numbers, not a published specification, and platforms adjust their interfaces without announcement. Treat them as a margin to design within rather than a boundary to sit exactly on.
Composing an on-model shot that survives every crop
The practical problem is that one shoot has to produce a 4:5 post and a 9:16 story. Those are different shapes, and cropping one into the other is where good images go to die: the 9:16 crop of a 4:5 frame usually removes the shoes, the head, or both.
- Shoot the tallest shape you needGenerate or shoot at 9:16 when a story is part of the plan, then crop down to 4:5. Cropping a wide frame up into a tall one always loses something.
- Keep the figure horizontally centredA centred figure survives a narrower crop from either side. A figure placed on a third looks better in one shape and gets cut in the other.
- Leave headroom and footroomSpace above the head and below the shoes is what a crop consumes. A frame composed tight in one ratio has nothing to give in another.
- Keep garment detail out of the cornersThe neckline, the hem, and any hardware should sit in the middle band. Detail in the corners is the first thing every crop and every interface element removes.
Text on the image, and where it goes wrong
Prices and badges are the most common thing to put on a fashion post and the most common thing to lose. Two failure modes account for nearly all of it: text placed in the bottom zone where the caption and buttons cover it, and text that was generated as part of an AI image and came back with a misspelling or a mangled numeral.
The first is a composition rule and costs nothing to follow. The second is structural: image models produce text as pixels rather than as characters, so spelling is probabilistic and every price has to be proofread. The way around it is to keep text out of the image model entirely and composite it onto the finished picture as a real text layer, which is what AURA does and why editing a price afterwards costs nothing and never regenerates the photograph.
Questions
- What size should an Instagram fashion post be?
- 1080 by 1350 pixels, a 4:5 portrait. It is the tallest shape a feed post allows, so it occupies the most screen while someone scrolls, which matters more for a full-length garment than for almost any other subject.
- Is 1:1 or 4:5 better for clothing?
- 4:5, in nearly every case. A square frame forces a choice between cutting a full-length look and shrinking the figure until fabric detail is lost. Use 1:1 only when the profile grid layout is the priority.
- What is the safe zone for a reel?
- On a 1080 by 1920 canvas, keep roughly the top 250 pixels and the bottom 420 pixels clear of anything important, and stay off the right edge where the action buttons sit. That leaves a reliably visible central band of about 1250 pixels of height.
- Why does text in AI-generated images come out misspelled?
- Because image models generate text as pixels rather than as characters, so the spelling is probabilistic. The reliable fix is to not send text through the image model at all and instead composite it onto the finished image as a real text layer.