Reprompt AI – Image to Prompt

Free image to prompt on Android · Google Play

Get
Reprompt.org

Text behind image: put text behind a photo

Upload a photo and this tool cuts the subject out in your browser, then sits your text between the background and that cutout — the text behind image look you see on Instagram, also called text behind photo or depth text. You can put text behind a person or an object, write more than one line, use your own fonts, add 3D perspective text, and give the type a border, a shadow or a gradient. The file you download is the same pixel size as the photo you started with, with no crop and no watermark. The picture never leaves your device.

Last updated

Drop a photo here, or choose one

The picture stays on your device. JPG, PNG or WebP. After the cutout, you can put text behind the subject, including 3D perspective text.

How do I put text behind an image?

Four steps, and only the cutout takes a moment. After that, every change redraws instantly, so you can try wording and placement before you download.

  1. Upload a photo. Choose a picture from your gallery, take one with the camera, or drop a file onto the page. JPG, PNG and WebP work in every modern browser.
  2. Wait for the subject cutout. A model runs in your browser and cuts the main subject — a person, a pet, or an object — away from the background. On portraits, a higher-quality MediaPipe person segmenter (with a hair class) is used; a general U-2-NetP model is the fallback for other subjects. The first run downloads the model, then it stays cached. The photo itself is never uploaded.
  3. Place and style your text. Type one or more lines, drag them into position, and resize or rotate. Set the font (or upload your own), size, weight, italic, underline, strikethrough, color, a gradient fill, opacity, letter spacing and line height. Add a stroke (border), a shadow, and 3D perspective text — tilt, skew and a depth/extrude shadow. Or tap a one-tap preset such as Bold Pop, Neon, Magazine, Outline, 3D Block or Minimal. A refine-edges slider and an erase/restore brush let you clean up the cutout.
  4. Pick a size and download a still, a GIF, or a video. Original keeps the photo’s pixel width and height, with no crop and no watermark. Or pick a size template — YouTube thumbnail 1280×720, Instagram 1080×1080 or 1080×1350, Story/Reels 1080×1920, or Pinterest 1000×1500 — then drag the frame to keep the subject centered. You can also export an animated text behind image GIF or a text behind image video (MP4 when the browser supports H.264, otherwise WebM). The text animates in behind the subject; the subject stays still.

For a slower walkthrough with tips on photos that work well, see the guide: how to put text behind an image (free, no app).

What is a text behind photo, or depth text, effect?

It is a layered text effect: the words sit in the scene, partly hidden by a person, a pet, or an object, instead of floating in a box on top. People also call it text behind object, depth text, or “put text behind a person.” It reads well on Instagram posts and story covers because the subject stays the focus and the type still has somewhere to live. This tool does that by cutting the subject out and drawing your text between the original photo and that cutout. You can also tilt, skew and extrude the type for 3D perspective text — the usual way to make 3D text for Instagram, Reels or a Story without leaving the browser.

How do I put text behind a person or an object?

Pick a photo with one clear subject and a little empty space around it — a portrait, a product, a car, a pet. After the cutout, drag the text so it crosses the subject. The subject stays in front, which is what makes the type look like it is in the scene, including when you put text behind a primary object rather than a person. If a line needs to sit fully on top, turn on “Bring text in front” for that layer. A stroke (border), a light shadow, or 3D perspective text help the words stay readable where they peek out from behind.

Can I use my own fonts and long text?

Yes. There is a set of Google Fonts, all under the SIL Open Font License, and you can upload a .ttf, .otf or .woff file of your own. Text can run over several lines, and you can add more than one layer. You can set size, weight, italic, underline, strikethrough, color, a gradient fill, opacity, letter spacing, line height, and rotation, then drag each layer into place. One-tap presets (Bold Pop, Neon, Magazine, Outline, 3D Block, Minimal, Story Title, Reels Hook) apply a finished style at once. That covers the usual gaps in phone apps: custom fonts, long text, and a border or shadow on the type.

What size is the exported file?

Original keeps the photo you uploaded. If you start with a 4032 × 3024 picture, the PNG or JPG is 4032 × 3024, with no crop, no “web quality” shrink, and no watermark. For JPG you pick the quality yourself. Size templates crop and fit to a post frame while keeping the subject centered — you can drag the frame — and they auto-size the type: a text behind image YouTube thumbnail at 1280×720, Instagram 1080×1080 or 1080×1350, Story/Reels 1080×1920, or Pinterest 1000×1500. Presets such as 3D Block, Magazine, Reels Hook and Story Title pair well with those frames.

Does my photo leave this device?

No. The cutout runs in your browser. On a portrait, a higher-quality MediaPipe person segmenter (Apache License 2.0, with a hair class, about 16 MB once) is used; a general U-2-NetP model (Apache License 2.0, about 4.5 MB) is the fallback for other subjects, or if you pick “Any object.” The photo itself is not sent to a server. You do not need an account. If a browser cannot open an iPhone HEIC file, save it as a JPG and try again.

How do I export an animated text behind image video or GIF?

After the cutout, choose an animation: rise from behind, fade and scale, typewriter, slide from the left or right, or a 3D flip. Set the duration (2–6 seconds) and preview it. The words move; the person or object stays still on top, so the depth effect holds in motion. Download a text behind image GIF, or a text behind image video — MP4 with H.264 when the browser can encode it, otherwise WebM. There is no watermark. Video is capped at 1080 pixels on the long side (720 for GIF) so the file stays shareable. Loop is optional on the GIF.

How do I make a text behind image YouTube thumbnail?

Tap the YouTube size template (1280×720). The tool fits your photo to that frame and keeps the subject in the middle; drag the frame if you want a different crop. Place the title so it sits behind the person or object, try 3D Block or Bold Pop if you want a thick hook, and export a PNG or JPG. The download is 1280×720, ready to use as a text behind image YouTube thumbnail. The same templates cover Instagram posts, Story and Reels frames, and Pinterest.

Can I add 3D perspective text and 3D text behind image effects?

Yes. Each text layer has tilt X, tilt Y, skew, and a depth/extrude shadow, so you can set 3D perspective text behind a person or a text behind object shot. The 3D Block and Reels Hook presets are a quick start. The warp is applied at the photo’s full resolution, so the export is not a cropped preview. This is a styling effect — it does not estimate a 3D scene from the photograph.

Text behind image FAQ

Is this text behind image tool free?

Yes. The tool is free, with no account and no watermark. It runs in your browser on this page.

Does my photo get uploaded to a server?

No. The picture stays on your device. The subject cutout runs locally, after a one-time download of the on-device models (MediaPipe’s person segmenter and, if needed, U-2-NetP). We do not receive the photo, and we do not ask you to log in.

Will the export be the same size as my photo?

If you leave Size on Original, the PNG or JPG uses the original pixel width and height, with no crop. You can also pick a size template and drag the frame: YouTube thumbnail 1280×720, Instagram 1080×1080 or 1080×1350, Story/Reels 1080×1920, or Pinterest 1000×1500. For JPG you can set the quality. There is no watermark. 3D text for Instagram still works on these frames.

Can I upload my own font and write long text?

Yes. You can use the bundled Google Fonts (all SIL Open Font License) or upload a .ttf, .otf or .woff file. Text can run over several lines, and you can add more than one text layer. Each layer can be italic, underlined or struck through, and can use a gradient fill.

Can I put a border, a shadow, or 3D perspective text?

Yes. Each layer has a stroke (width and color), a shadow (color, blur and offset), and 3D perspective controls — tilt X, tilt Y, skew, and a depth/extrude shadow. That is often what makes layered text effects readable when they sit behind a person or an object, and it is how you get 3D text behind image looks for a post or a Reel.

How do I put text behind a person or an object?

Upload a photo with a clear subject and a little empty space around it. After the cutout finishes, drag the text so it crosses the subject — the subject stays in front, which is the text-behind-photo (or depth text) look. This is also how you put text behind a primary object in a product or pet photo. If you want the type fully visible, turn on “Bring text in front” for that layer.

Can I export an animated text behind image video or GIF?

Yes. Pick an animation — rise from behind, fade and scale, typewriter, slide from the left or right, or a 3D flip — set a duration of 2 to 6 seconds, and preview it on the photo. Download a text behind image GIF (always available) or a text behind image video. Video uses MP4 (H.264) when your browser supports it, otherwise WebM. The text animates in; the subject stays on top. There is no watermark. To keep files light, video is capped at 1080 pixels on the long side, and GIF at 720.

How do I make a text behind image YouTube thumbnail?

Upload your photo, then tap the YouTube size (1280×720). The frame is a smart crop that keeps the subject centered — drag it if you want a different slice. Place your text so it sits behind the person or object, try the 3D Block preset if you want a thick title, and download a PNG or JPG at 1280×720, ready for a text behind image YouTube thumbnail.

When does the cutout struggle, and can I refine it?

It works best with one clear subject. On people, the MediaPipe path is usually stronger on hair than the general model; on pets, cars and products, the general U-2-NetP path is the better default (Auto picks this for you). Thin details, glasses, or a subject that blends into the background can still leave rough edges. Use the refine-edges sliders (threshold and feather) and the erase/restore brush to clean those up. You can also switch the cutout path by hand.

Want a prompt from a picture instead of a text effect? Try the image to prompt tool. For a prank photo with a ghost sitting behind someone, use the ghost prank photo editor — same on-device cutout, a drawn ghost instead of type.

Get weekly AI insights, free

One email a week with AI news and insights from the Era of Minds newsletter.

Free. No spam, unsubscribe any time. By subscribing you agree to receive emails from Era of Minds. Form not showing? Subscribe on beehiiv.