Blog

Practical guides on prompts, captions, hashtags and writing about images — for creators, marketers and designers who'd rather make things than read 4,000-word filler posts.

Writing6 min read

How to Describe an Image (Without Sounding Like a Stock Photo Caption)

A simple, repeatable framework for writing image descriptions that feel human — useful for alt text, captions, and AI prompts.

Social5 min read

Best AI Captions for Instagram: What Actually Stops the Scroll

What makes a caption land in 2026 — and how to use AI to draft them faster without sounding like a chatbot.

Prompts7 min read

AI Image Prompt Examples: A Cheat-Sheet by Style

Real, copy-pasteable prompts broken down by subject, style and mood — and the simple structure underneath them.

Social5 min read

How to Generate Better Hashtags (and Why Walls of 30 Are Over)

A modern hashtag strategy: smaller, smarter sets that mix specific, category and aesthetic tags.

Inspiration4 min read

Interior Design Caption Ideas: 30 Lines You Can Steal

Caption starters for interiors, mood boards, styled rooms and before/afters — organised by tone.

Read in full
Writing

How to Describe an Image (Without Sounding Like a Stock Photo Caption)

A simple, repeatable framework for writing image descriptions that feel human — useful for alt text, captions, and AI prompts.

Most image descriptions read like they were written by a tired intern at a stock photo agency. "A woman holding a cup of coffee." Technically correct. Completely forgettable. The good news: there's a simple framework you can use to write descriptions that are accurate, vivid, and genuinely useful — whether you're writing alt text, an Instagram caption, or a prompt for an AI image generator.

Start with the subject, not the setting

Lead with what (or who) the image is about. "A potter shaping a tall stoneware vase" is stronger than "A studio with shelves of pottery, and a person at a wheel." Setting is context — it belongs in the second half of the sentence, not the first.

Add one specific, sensory detail

Specificity is what separates a description from a label. Instead of "a coffee shop," write "a small coffee shop with steam fogging the window." Instead of "a sunset," write "a sunset the color of unripe peaches." One precise detail outperforms three vague ones.

Name the mood

Every image has a tone — calm, chaotic, intimate, expansive, melancholy, playful. Naming it out loud (even just one adjective) makes the description feel intentional rather than mechanical.

Mention composition only if it matters

If the framing is part of the message — a tight close-up, an overhead flat-lay, a wide landscape with the subject pushed to the edge — say so. If it's a straightforward shot, skip it. Composition notes are signal, not filler.

Cut everything that isn't doing work

Read your description out loud. Delete the words that don't add information, mood, or rhythm. "A beautiful image of a really nice…" is almost always cut-able. Short sentences land harder.

A worked example

Before: "A picture of a dog running on the beach during sunset." After: "A wet golden retriever sprinting along the tide line at dusk, ears flat, sand kicking up behind her, the sky a soft coral fade." Same image. One reads like metadata. The other puts you there.

When in doubt, let AI draft and you edit

Drafting from scratch is the slow part. Tools like DescribeThis.ai generate a usable first pass in a few seconds — caption, description, mood, even hashtags. Treat the output as a draft, apply the rules above, and you'll have something publishable in under a minute.

Try it on one of your own images — drop a photo into DescribeThis.ai and you'll get a prompt, a caption, a vibe and hashtags in a few seconds.

Social

Best AI Captions for Instagram: What Actually Stops the Scroll

What makes a caption land in 2026 — and how to use AI to draft them faster without sounding like a chatbot.

The Instagram caption has quietly become one of the hardest pieces of micro-copy on the internet. You have one or two lines visible before the "more" cut-off. The algorithm rewards saves and shares more than likes. And generic AI-generated captions are increasingly easy to spot.

What a good caption does

It earns the second line. The first line either creates curiosity, lands a clean joke, says something specific enough to feel honest, or names a feeling people recognise. Everything else is a bonus.

The three caption shapes that consistently work

1) The observation. A short, specific noticing about the moment in the photo. ("Slow mornings, warm cups, and absolutely no agenda.") 2) The contrast. Two things in tension. ("Looks expensive. Costs nothing.") 3) The invitation. A question or prompt that invites a comment. ("What's the last place that made you feel like this?")

How to use AI without sounding like one

Generic AI captions all share the same fingerprints: vague positivity, three emoji, a stock motivational line. Avoid them by giving the AI a strong starting point — an actual image, a tone ("dry," "warm," "deadpan"), and a length constraint ("under 12 words"). Then edit. The fastest way to a caption that sounds like you is to AI-draft three options and rewrite the best one in your own voice.

On hashtags

Hashtag strategy in 2026 leans small. Three to seven well-chosen tags — a mix of specific (what's in the image), category (the broader topic), and aesthetic (the mood) — outperform walls of 30. We cover this in detail in our hashtag guide.

The shortcut

DescribeThis.ai writes captions in the shape above, plus a curated hashtag mix, from any image. Use it as your first draft engine, then make it sound like you.

Have an image you'd like to start from? Upload it to DescribeThis.ai and let it draft the prompt, caption and hashtags — then edit them in your own voice.

Prompts

AI Image Prompt Examples: A Cheat-Sheet by Style

Real, copy-pasteable prompts broken down by subject, style and mood — and the simple structure underneath them.

Most "prompt guides" are 4,000 words of theory and zero usable prompts. This isn't that. Below is a cheat-sheet of prompts you can lift directly, plus the underlying structure so you can adapt them to whatever you're making.

The prompt structure that works almost everywhere

Subject + style + lighting + composition + medium + mood. Not every prompt needs all six, but if a prompt is underperforming, it's usually missing one of them. The order matters less than the presence.

Editorial photography

"A young chef plating microgreens in a sunlit Copenhagen kitchen, shot on 35mm film, shallow depth of field, soft window light from the left, overhead three-quarter angle, warm muted palette, quiet and contemplative mood."

Cinematic landscape

"A lone hiker on a basalt ridge above a sea of clouds at dawn, anamorphic widescreen, low golden sun, rule-of-thirds composition with the figure on the right, painterly haze, Roger Deakins lighting, awestruck and serene."

Product / brand

"A matte ceramic perfume bottle on a textured travertine plinth, soft three-point studio lighting, hero centred composition, shallow depth of field, dust motes catching the back light, minimal Scandinavian palette of bone and clay."

Illustrated / editorial

"Flat editorial illustration of a person reading in a window seat at night, limited four-colour palette of navy, terracotta, cream and ink, slightly grainy texture, geometric shapes, New Yorker-style line work, calm and introspective."

3D / render

"Stylised isometric 3D render of a tiny rooftop garden, Octane render, soft ambient occlusion, pastel palette, slight depth-of-field blur on background, playful and inviting."

Where DescribeThis.ai fits

If you already have a reference image — a photo, a screenshot, a mood-board pull — drop it into DescribeThis.ai and you'll get a prompt in this structure, automatically. From there it's a 10-second edit instead of a 10-minute write.

The fastest way to apply any of this is to start from the image itself. DescribeThis.ai will give you a clean first pass to refine in under a minute.

Social

How to Generate Better Hashtags (and Why Walls of 30 Are Over)

A modern hashtag strategy: smaller, smarter sets that mix specific, category and aesthetic tags.

The era of 30 hashtags stuffed into the first comment is over. Most platforms now reward small, relevant, well-chosen hashtag sets — and penalise the spammy ones. Here's a strategy that works across Instagram, Threads, TikTok and even Pinterest.

The 3-3-3 mix

Aim for roughly three of each: three specific tags (what's literally in the image — #ceramicstudio, #goldenretriever, #osakatemple), three category tags (the broader bucket — #pottery, #dogsofinstagram, #travelphotography), and three aesthetic or mood tags (#slowliving, #goldenhour, #warmtones). Nine focused tags consistently outperform twenty-five vague ones.

Match the size of the tag to the size of your account

If your account is small, mega-tags like #love or #photography won't get you discovered — you'll drown. Mix in medium and small tags (10k–500k posts) where your content has a chance of staying near the top. Tools like DescribeThis.ai suggest tag mixes calibrated to the image, which gives you a useful starting point you can prune.

Avoid these common mistakes

• Hashtag stuffing ("#love #life #happy #good #nice") • Banned or shadow-banned tags (do a quick search before using) • Tags that don't match the image (it hurts reach more than it helps) • Recycling the exact same hashtag block on every post

A practical workflow

1) Generate a starter set from your image (DescribeThis.ai does this in a click). 2) Cut anything generic or unrelated. 3) Swap one or two for tags specific to your niche or community. 4) Save your edited set as a note so you can reuse and iterate.

Try it on one of your own images — drop a photo into DescribeThis.ai and you'll get a prompt, a caption, a vibe and hashtags in a few seconds.

Inspiration

Interior Design Caption Ideas: 30 Lines You Can Steal

Caption starters for interiors, mood boards, styled rooms and before/afters — organised by tone.

Interior shots are some of the hardest to caption. The image is usually beautiful but quiet, and "new sofa!" doesn't quite cut it. Here are 30 caption starters, organised by tone, that work for designers, stylists, real estate agents, and weekend renovators.

Warm and lived-in

• "Rooms that feel like a long exhale." • "This corner does more for me than therapy." • "Soft light, warm wood, no agenda." • "The kind of room that asks you to stay a little longer." • "Built for slow Sundays."

Confident and editorial

• "Restraint as a design choice." • "One material, three textures, zero noise." • "A small room that thinks like a big one." • "Quiet luxury, said quietly." • "Proof that beige can have a personality."

Playful

• "Couch goals. That's the whole caption." • "Romanticising the entryway." • "This rug walked so the rest of the room could run." • "My ick-list is short and this room is on the safe side of it." • "Built for vibes. Furnished by accident."

Process / before & after

• "Six weekends, four delivery delays, one perfect light switch." • "Before, during, and 'please don't ask about the budget.'" • "Every project teaches you one thing. This one taught me three." • "From beige rental to actually-mine." • "Small change, disproportionate joy."

Mood board / inspiration

• "Saving this for the next project, the next house, and the next life." • "References, not rules." • "Mood: warm minimalism with one ugly chair." • "If I could live inside a colour palette, it would be this one." • "Tomorrow's mistakes start here."

Pair with the image

Run any interior shot through DescribeThis.ai and you'll get a custom caption, a vibe summary and a hashtag mix calibrated to that specific room — a much faster starting point than scrolling through caption lists.

Have an image you'd like to start from? Upload it to DescribeThis.ai and let it draft the prompt, caption and hashtags — then edit them in your own voice.