Add Text to Image

What people add text to images for

Most of the jobs are small and practical: labelling a screenshot for a how-to, captioning a photo before sending it to a group, putting a date and place on a picture, marking up a product photo, making a quote card or an announcement for social media, adding a title to a video thumbnail or a slide image. None of them justify installing a design application or signing up for an online one.

A blank canvas covers the graphics that have no photo at all — a coloured square with a quote, a story-sized announcement, a simple title card. Pick the size and colour, then add text exactly as you would to a photo.

Making text readable on a photograph

Text on a photo fails for one reason: not enough contrast between the letters and whatever is behind them, which in a photograph changes from one letter to the next. There are four ways to fix it, in increasing order of strength.

Placement comes first. Put text over a calm area — sky, a wall, an out-of-focus background — rather than over detail. A drop shadow helps next: a soft dark shadow behind light text separates it from mid-tones without being obvious. An outline is stronger, giving each letter a hard edge in a contrasting colour, and is the reason meme captions read on any picture. Strongest of all is a background box: a solid or semi-transparent panel behind the text, which guarantees legibility at the cost of covering some of the picture.

As a rule of thumb from web accessibility guidance, body-size text needs a contrast ratio of at least 4.5 to 1 against its background, and large text 3 to 1. You cannot measure that across a busy photograph, but a background box at 60–80% opacity in a dark colour behind white text clears it comfortably. The colour palette tool on this site can check exact ratios between two colours.

Fonts, sizes and how the text is drawn

The font list uses typefaces already installed on your device: sans-serif (Arial, Helvetica), the system interface font, serif (Georgia, Times), monospace, condensed, rounded, Comic Sans, a script face, and Impact for meme-style captions. Nothing is downloaded, which keeps the page fast and private, but it does mean the exact face differs between devices — a Mac, a Windows PC and an Android phone each have their own. Each choice is a stack of similar fonts, so you always get something close.

Once saved, the font no longer matters. The text is drawn into the image as pixels, so anyone who opens the file sees exactly what you saw, with no font of their own needed. The flip side is that the text cannot be edited later in another program — keep the original picture if you might want to change the wording.

Size is set as a percentage of the image's width rather than in pixels, so 5% is 54 pixels tall on a 1080-pixel post and 200 pixels on a 4000-pixel photo. That keeps the preview and the full-resolution file identical. Line width controls wrapping; a long sentence wraps onto several lines within it, and alignment sets whether those lines are flush left, centred or flush right.

Positioning precisely

Drag a box on the preview to move it. For finer placement, select the box and use the arrow keys (one percent of the image per press, five with Shift) or the horizontal and vertical position sliders, which also make the tool usable without a pointer. Rotation turns a box around its own centre, for a diagonal label or a stamp effect.

Common layouts: a title in the top third with a quieter subtitle beneath; a caption in the bottom quarter, centred, on a background box; a small credit or date in a corner at 2–3% size. Leave a margin from the edges — social platforms crop and round corners, and text pressed against the edge looks accidental.

Edits apply to the selected box only, unless you tick Apply style to every box, which is useful when several labels should match.

Saving, and the limits

The image is saved at its original resolution, not at the size of the preview. PNG keeps the edges of letters perfectly sharp and is the right choice for graphics and screenshots; JPEG is far smaller for photographs and keeps text acceptably sharp at quality 90 and above. Very large photos are scaled to the biggest canvas your browser allows, and the page says so if that happens.

Scripts that browsers render well — Devanagari, Arabic, Cyrillic, Chinese, Japanese and Korean among them — work here too, as long as your device has a font for them. Text without spaces between words, such as Chinese or Japanese, wraps between characters. Emoji are drawn in your system's emoji style.

What it does not do: custom font uploads, text on a curve, letter-spacing control, or editing text after the image is saved. The picture is processed on your device and never uploaded.