How to Name Images for SEO

TL;DR — Name your image files something that describes what’s in the picture, using real words separated by hyphens: golden-retriever-puppy.jpg, not IMG_0042.jpg. Do this for new images. Don’t go back and rename all your old ones — that causes more headaches than it’s worth, for a benefit you’ll never notice.

What an image file name is (and why it matters a little)

Every image on your site has a file name — the bit at the end of its URL, like sunset.jpg or IMG00023.JPG. A search engine can’t “see” your photo the way you do, so it looks for clues about what the image shows. The file name is one of those clues.

It’s a small clue. Google leans much more on your alt textAlt text is the value of the `alt` attribute on an HTML `<img>` element — a short text substitute chosen for the image's purpose and context, not a literal description of what it shows. It makes images accessible to screen-reader users and helps search engines understand images, mainly for image search. and the words around the image than on the file name. But a descriptive name is free to get right on new images, and it mostly helps you show up in image search (Google Images, Google Lens) rather than regular web results.

The naming rules

• Describe the image. black-kitten.jpg tells Google (and anyone who sees the URL) what the picture is. IMG00023.JPG or image1.jpg tells them nothing.
• Use hyphens between words. Write red-running-shoes.jpg, not red_running_shoes.jpg or redrunningshoes.jpg. Google treats a hyphen as a space between words; an underscore glues the words together.
• Keep it lowercase and short. A few descriptive words is plenty. You don’t need a sentence.
• Don’t stuff keywords. shoes-buy-shoes-cheap-shoes-red-shoes.jpg looks spammy and doesn’t help. One honest description is the goal.

The thing most people get wrong

Don’t rename images you’ve already published. It’s tempting to go back and give your whole image library “better” names. Don’t. Renaming a file changes its web address, which means broken links, redirectsA redirect sends browsers and crawlers from a requested URL to a different one. An HTTP redirect specifically is a 3xx status code paired with a Location header; meta refresh and JavaScript redirects achieve a similar navigation without being a 3xx response themselves. Permanent redirects (301/308) are Google's signal the target should be canonical; temporary ones (302/303/307) aren't. to set up, and weeks (sometimes months) of your images looking wrong in Google Images while Google slowly notices the change. Google itself has said the payoff is basically nothing if your alt textAlt text is the value of the `alt` attribute on an HTML `<img>` element — a short text substitute chosen for the image's purpose and context, not a literal description of what it shows. It makes images accessible to screen-reader users and helps search engines understand images, mainly for image search. is already decent. Get the name right on new images and leave the old ones alone.

One more thing worth knowing: the file extension (.jpg, .png, .webp) is just a label you gave the file — it doesn’t guarantee what format it actually is. That’s a technical detail, not something to worry about day to day, but it matters if you’re ever converting or migrating images.

This piece sits under my Image SEOImage SEO is optimizing the images on your pages so search engines can discover, crawl, index, and rank them — in Google Images and visual search, and as part of standard web results. It spans file format, filenames, alt text, compression, responsive markup, structured data, and image sitemaps. hub — for how file names fit alongside alt text, formats, and page speed, start there. Want the exact Google quotes, the hyphens-vs- underscores mechanics, and the redirect/cache details on renaming? Switch to the Advanced tab.

TL;DR — Image file names are a real but light signal for image search — Google reads them alongside alt textAlt text is the value of the `alt` attribute on an HTML `<img>` element — a short text substitute chosen for the image's purpose and context, not a literal description of what it shows. It makes images accessible to screen-reader users and helps search engines understand images, mainly for image search., surrounding text, and the image itself, not as a standalone ranking lever. Best practice: short, descriptive, hyphenated, lowercase names on new images (Google’s canonical example is my-new-black-kitten.jpg over IMG00023.JPG). The underused, high-value point: don’t bulk-rename already-indexedStoring a crawled page in the search index so it can appear in results. Crawled is not the same as indexed — Google selects what to keep, and indexing isn't guaranteed. images. Mueller warned it takes Google months to re-see them and the effect is “minimal effect, maybe no visible effect at all” if alt textAlt text is the value of the `alt` attribute on an HTML `<img>` element — a short text substitute chosen for the image's purpose and context, not a literal description of what it shows. It makes images accessible to screen-reader users and helps search engines understand images, mainly for image search. and context are already good — and because renaming changes the URL, you inherit 301s, 404 risk, and CDN/browser cache churn for an unmeasurable gain. File names complement alt text; they don’t replace it.

What “image file name SEO” actually means

Start with the boundary, not the technique: Google’s own framing calls the filename a “very light” clue, and nothing here promises a ranking lift — a perfect file name is a small hygiene win, not a lever you pull for traffic. Within that boundary, naming an image file for SEO means giving it a name that describes its subject in real words — golden-retriever-puppy.jpg instead of IMG00023.JPG — so the name itself is a usable clue about what the image shows. Google reads that name as one of several weak signals, alongside alt text, the text surrounding the image, and the pixels themselves, to work out what an image depicts. The payoff is mostly in Google Images and Lens ranking, not general web-search ranking.

Keep the scale honest: this is a light signal. My Image SEOImage SEO is optimizing the images on your pages so search engines can discover, crawl, index, and rank them — in Google Images and visual search, and as part of standard web results. It spans file format, filenames, alt text, compression, responsive markup, structured data, and image sitemaps. hub frames the whole topic as two separate jobs — ranking in image search, and keeping pages fast — and file names sit firmly on the image-search side, well below alt text in importance. A perfect file name won’t rescue an image with no alt text and no relevant page around it.

The core naming rules

Descriptive over generic

Google’s Image SEO best practices put the rule plainly: “When possible, use filenames that are short, but descriptive. For example, my-new-black-kitten.jpg is better than IMG00023.JPG.” And it’s explicit about the failure mode: “Avoid using generic filenames like image1.jpg, pic.gif, 1.jpg when possible.” The reasoning is that “the filename can give Google very light clues about the subject matter of the image” — “very light,” but not zero, and a generic name throws that clue away for free.

Hyphens vs. underscores

Use hyphens to separate words: black-kitten.jpg, not black_kitten.jpg. The mechanics are worth understanding rather than repeating as folklore. A hyphen is read as a word boundary — Google parses black-kitten as the two words “black” and “kitten.” An underscore joins the tokensA token is the smallest unit of text (or image/audio/video) an LLM processes — roughly 4 characters, or about ¾ of an English word. A context window is the maximum number of tokens (input plus output) a model can hold at once, like its short-term memory. — black_kitten is more likely to be read as the single string “black_kitten.” So red-running-shoes.jpg communicates three words; red_running_shoes.jpg communicates something closer to one.

A caveat on provenance: Google’s Image SEOImage SEO is optimizing the images on your pages so search engines can discover, crawl, index, and rank them — in Google Images and visual search, and as part of standard web results. It spans file format, filenames, alt text, compression, responsive markup, structured data, and image sitemaps. page doesn’t spell out “hyphens, not underscores” in those literal words for filenames specifically. The direct source is broader: Google’s URL structure guidance recommends hyphens over underscores as word separators across URLs generally, not as an image-specific rule with its own measured effect size. Older Matt Cutts-era commentary on the same word-separator behavior is where the SEO-blog folklore version of this rule comes from, and it’s now near-universal practice. Treat it as inherited, URL-wide convention rather than a filename-specific Google quote — but the underlying mechanism (hyphen = boundary, underscore = joined) is accurate and safe to follow. You won’t be penalized for underscores; hyphens just communicate more.

Lowercase and short

Keep names lowercase and roughly three to eight words as a practical rule of thumb — Google doesn’t publish a fixed ideal word count, so treat that range as sensible convention, not a measured threshold. Some servers treat Photo.JPG and photo.jpg as different files (case-sensitive paths can resolve to two distinct URLs, not one), so lowercasing avoids that surprise. Short-but-descriptive beats long-and-exhaustive.

Keyword use without stuffing

Include the words that genuinely describe the image — that’s the point. What to avoid is cramming: red-shoes-buy-red-shoes-cheap-red-shoes.jpg reads as spam and helps nothing. One nuance to be precise about: Google’s explicit keyword-stuffing warning is written against alt text, not file names — “Avoid filling alt attributes with keywords (also known as keyword stuffing) as it results in a negative user experience.” The industry reasonably extends the same “don’t overdo it” logic to file names, but don’t misattribute the warning: Google’s literal line targets the alt attribute.

Localize file names on translated sites

If you run localized versions of a page, Google suggests matching the file names: “If you localize your images, remember to also translate the filenames, keeping in mind the URL encoding guidelines.” Because the file name is part of the URL, non-Latin or accented characters need proper URL encoding.

The extension isn’t proof of the actual format

A file name’s extension is a label, not a guarantee. .jpg, .png, and .webp are what you called the file — they don’t prove what bytes are actually inside it or what format a browser or crawlerA crawler — also called a spider or bot — is an automated program that fetches web pages, extracts their links, and queues new URLs to visit. Search engines use crawlers to discover and download content for their index. will treat it as. Two other signals matter more:

• The Content-Type response header is what the server declares the file to be when it’s requested. It should match the actual encoded bytes — MDN’s Content-Type reference documents this declaration, and notes that when a server sends X-Content-Type-Options: nosniff, a mismatched or missing Content-Type can cause the resource to be blocked outright rather than guessed at.
• The decoded bytes themselves are the ground truth. Browsers fall back to sniffing the actual byte signature when a Content-Type is missing or generic — the MIME Sniffing Standard is the spec that defines how a “supplied” type (extension, header) and a “computed” type (what the bytes actually are) can diverge, and which one wins in practice.

Practical takeaway: if you convert an image (say, .png to .webp for compressionCompression (HTTP content encoding) shrinks text-based responses — HTML, CSS, JS, JSON, SVG, XML sitemaps — before they're sent over the network, using an algorithm like Gzip, Brotli, or Zstd, so the browser or crawler downloads fewer bytes. It's not a ranking factor, but it speeds up page loads and helps pages stay under crawler fetch limits.) but don’t rename the file or update the server’s declared type, you can end up with a .png-named file whose header and bytes say webp, or vice versa. That’s a real, checkable mismatch — not an SEO ranking factor on its own, but a technical-hygiene item worth auditing alongside naming. (Format choice and compressionCompression (HTTP content encoding) shrinks text-based responses — HTML, CSS, JS, JSON, SVG, XML sitemaps — before they're sent over the network, using an algorithm like Gzip, Brotli, or Zstd, so the browser or crawler downloads fewer bytes. It's not a ranking factor, but it speeds up page loads and helps pages stay under crawler fetch limits. themselves are covered in my Image FormatsThe file encoding used to store an image (JPEG, PNG, WebP, AVIF, SVG, GIF) — it drives file size and Core Web Vitals, but carries no direct ranking weight. and Image Optimization guides, not here.)

Should you rename images that are already indexed?

This is the question nearly every naming guide skips, and it’s where the real money is — because the honest answer is usually no.

What Mueller and Sassman said

On Google’s Search Off the Record podcast (recapped by Search Engine Journal), Lizzi Sassman and John Mueller worked through exactly this. Mueller confirmed the signal is real but minor: descriptive file names are recommended, but “I don’t think you would see a significant change if you already do the other things around images, like the alt texts, the text surrounding the image.” His verdict on the upside of a rename: “That probably would have minimal effect, maybe no visible effect at all.”

The bigger warning was about doing it in bulk. Google re-crawls images infrequently — “when we crawl images, we tend not to crawl them as often, because usually, they don’t change a lot” — so a mass rename lingers. Mueller’s own estimate was hedged even in the original conversation: “if you changed all of them at once, my guess is… I don’t know, over a period of a couple of months at least, it’ll be kind of annoying in Image Search.” Treat “a couple of months” as Mueller’s rough, in-the-moment guess rather than a measured figure — it’s reported via SEJ’s recap of the podcast, not a primary transcript I’ve independently re-verified, so don’t repeat it as a precise recovery window. The direction (slow, multi-week-to-multi-month disruption) is the durable takeaway; the exact number isn’t. Sassman flagged the operational side too — “room for human error… to miss a broken link” when you’re hunting down everywhere an image is embedded.

Redirects, 404s, and cache implications

Here’s the mechanical reason bulk renaming is costly, and it’s the piece competitor guides leave out. Renaming an image file changes its URL. That has knock-on effects:

• The old URL now 404s unless you add a 301 redirectA 301 redirect is the HTTP status code for a permanent move: it tells browsers and search engines a URL has moved for good, and it's the strongest signal for consolidating a page's ranking signals onto the new URL. Google says permanent redirects don't cause a loss in PageRank. from old path to new.
• CDN and browser caches may keep serving the old path (or cachingCaching stores a copy of a page or resource — in a browser, a CDN edge node, or a search crawler's own cache — so it can be served again without regenerating or re-downloading it. It isn't a direct ranking factor, but it feeds page speed and crawl efficiency. the new 404) for a while, so the change doesn’t propagate cleanly.
• Google re-crawls images slowly (per Mueller above), so the “before” version can sit in Google Images for weeks to months after you’ve moved on.

Gary Illyes framed the general signal loss from image URL migrations as “in line with web search results, which is a few weeks” — that’s about migrating image URLs broadly, not renaming specifically, but it’s the same category of temporary disruption. Put the pieces together and a site-wide rename is real engineering work and weeks of Image Search churn in exchange for an effect Mueller called “maybe no visible effect at all.”

When a rename is actually worth it

Not never — just not as a standalone SEO project. Rename opportunistically when you’re already touching the file: a page rebuild, a CMSA content management system (CMS) is software that lets users create, manage, and publish digital content — like blog posts and pages — without writing raw code. WordPress, Drupal, and Joomla are the most common open-source CMS platforms. migration, a redesign that’s changing image URLs anyway. At that point the redirectA redirect sends browsers and crawlers from a requested URL to a different one. An HTTP redirect specifically is a 3xx status code paired with a Location header; meta refresh and JavaScript redirects achieve a similar navigation without being a 3xx response themselves. Permanent redirects (301/308) are Google's signal the target should be canonical; temporary ones (302/303/307) aren't. and cache work is already on your plate, so fixing a bad file name is nearly free. The mistake is spinning up a dedicated “rename everything for SEO” initiative for marginal, unmeasurable upside.

Treat every rename as a URL migration, and audit it like one

A renamed image file is a new URL. Whenever you do rename — opportunistically or otherwise — run through the same checklist you’d use for any URL migration, not just a naming checklist:

• New name quality: descriptive, hyphenated, lowercase — the rules above.
• Reachability: does the new URL actually resolve (200, not a typo’d path)?
• Redirect: does the old URL 301 to the new one, so nothing 404s?
• References: did every internal linkAn internal link is a hyperlink from one page on a website to another page on the same website. Internal links help search engines discover your pages and pass ranking signals (PageRank and anchor-text context) between them., sitemapA sitemap is a file that lists the pages, images, videos, and other files on your site so search engines can discover them. It helps discovery, but submitting a sitemap doesn't guarantee crawling or indexing. entry, and CMS field that pointed at the old path get updated to the new one, or are you relying on the redirect to paper over stragglers?
• Declared vs. actual type: does the server’s Content-Type for the new URL match the real decoded bytes (see “The extension isn’t proof of the actual format” above)? A rename is also a reasonable moment to catch a stale mismatch you inherited.

Naming, format, and compression are separate jobs — this checklist covers naming and URL stability only. Format and codec choices belong in Image FormatsThe file encoding used to store an image (JPEG, PNG, WebP, AVIF, SVG, GIF) — it drives file size and Core Web Vitals, but carries no direct ranking weight.; compression and delivery belong in Image Optimization.

You probably can’t measure it

Worth saying plainly: even when you do rename, expect not to be able to prove it did anything. Mueller’s “minimal, maybe no visible effect at all” cuts both ways — if the gain is invisible in the SERPs, it’ll be invisible in your analytics too. Don’t build a business case around a file-name change.

How file names relate to alt text

File names and alt text are related but not interchangeable, and it’s worth keeping them straight:

• Alt text is the stronger, primary signal for image search and is required for accessibility. The file name is a much lighter clue.
• Screen readers fall back to reading the raw file name aloud only when alt text is missing or broken — which is a real accessibility argument for descriptive names, but a fallback, not the main event. (Full treatment in my alt text guideAlt text is the value of the `alt` attribute on an HTML `<img>` element — a short text substitute chosen for the image's purpose and context, not a literal description of what it shows. It makes images accessible to screen-reader users and helps search engines understand images, mainly for image search..)

So a good file name is not a substitute for alt text; it’s a complementary, weaker signal that happens to also help the accessibility fallback case. Do both — write proper alt text and name the file descriptively — rather than treating one as covering for the other.

Before / after examples

Common mistakes to avoid

• Bulk-renaming an existing, indexed image library for a marginal SEO gain — weeks of Image Search disruption plus redirect/404/cache work.
• Renaming without adding a 301, leaving the old URL to 404.
• Underscores instead of hyphens — not penalized, but weaker word parsing.
• Generic names (image1.jpg, pic.gif, 1.jpg, DSC_0491.jpg).
• Keyword-stuffed names that read as spam.
• Treating the file name as a replacement for alt text — it’s the lighter of the two signals and doesn’t cover accessibility.
• Forgetting to translate/encode file names on localized pages.
• Trusting the extension as proof of the real format. Why it’s wrong: .jpg, .png, and .webp are labels, not guarantees — the server’s Content-Type header and the actual decoded bytes are what a browser or crawlerA crawler — also called a spider or bot — is an automated program that fetches web pages, extracts their links, and queues new URLs to visit. Search engines use crawlers to discover and download content for their index. actually uses, and a format conversion that skips updating one of those creates a real, checkable mismatch. What to do instead: when you convert or rename a file, check that the extension, declared Content-Type, and decoded bytes all agree.

FAQs

Does the image file name affect SEO? Slightly, and mostly for image search. Google reads it as a very light clue alongside alt text, surrounding text, and the image itself. It’s not a standalone ranking lever.

Hyphens or underscores? Hyphens. Google reads a hyphen as a word boundary and an underscore as joining words into one token, so hyphens communicate more.

Is it bad to rename images after they’re indexed? In bulk, usually yes — Mueller warned it can be “annoying in Image Search” for a couple of months, and you take on redirect/404/cache work for an effect he called “minimal, maybe no visible effect at all.” Rename opportunistically during a rebuild, not as a standalone project.

How many words should a file name have? Roughly three to eight — short but descriptive. Enough to describe the subject, not a sentence.

Can I put keywords in the file name? Yes, if they genuinely describe the image. Don’t stuff. Google’s explicit keyword-stuffing warning targets alt text, but the same “don’t overdo it” logic applies.

Does the file name matter if I already have good alt text? Barely. Mueller: you won’t see a significant change from file names if your alt text and surrounding context are already good. Do file names right on new images; don’t expect them to move the needle on their own.

What happens to Google Images rankings if I rename my files in bulk? Temporary disruption — the old versions linger for a stretch (Mueller’s own estimate, “a couple of months,” is a rough guess from a recap, not a measured figure) because Google re-crawls images slowly, and you’re managing redirects and caches the whole time.

Does the file extension prove what format an image actually is? No. .jpg, .png, and .webp are labels you chose, not a guarantee of the encoded bytes. The server’s Content-Type header and the actual decoded bytes are what browsers and crawlers rely on — treat the extension as a naming convention, and check Content-Type and bytes separately if you’re auditing a conversion or migration.

AI summary

A condensed take on the Advanced version:

• Image file names are a real but light signal, mostly for image search (Google Images / Lens). Google reads the name alongside alt textAlt text is the value of the `alt` attribute on an HTML `<img>` element — a short text substitute chosen for the image's purpose and context, not a literal description of what it shows. It makes images accessible to screen-reader users and helps search engines understand images, mainly for image search., surrounding text, and the image itself — it’s not a standalone ranking lever.
• Naming rules: descriptive over generic (my-new-black-kitten.jpg beats IMG00023.JPG); hyphens, not underscores (hyphen = word boundary, underscore = joined tokenA token is the smallest unit of text (or image/audio/video) an LLM processes — roughly 4 characters, or about ¾ of an English word. A context window is the maximum number of tokens (input plus output) a model can hold at once, like its short-term memory. — sourced from Google’s general URL-structure guidance, not an image-specific rule); lowercase and short (≈3–8 words as convention, not a Google- measured threshold); include real keywords but don’t stuff.
• Keyword-stuffing nuance: Google’s explicit stuffing warning targets alt textAlt text is the value of the `alt` attribute on an HTML `<img>` element — a short text substitute chosen for the image's purpose and context, not a literal description of what it shows. It makes images accessible to screen-reader users and helps search engines understand images, mainly for image search., not file names — the “don’t overdo it” logic extends by convention, but don’t misattribute the quote.
• Extension ≠ actual format. .jpg/.png/.webp are labels; the server’s Content-Type header and the decoded bytes are the real signals a browser or crawlerA crawler — also called a spider or bot — is an automated program that fetches web pages, extracts their links, and queues new URLs to visit. Search engines use crawlers to discover and download content for their index. uses. A conversion that doesn’t update all three creates a checkable mismatch (format/compressionCompression (HTTP content encoding) shrinks text-based responses — HTML, CSS, JS, JSON, SVG, XML sitemaps — before they're sent over the network, using an algorithm like Gzip, Brotli, or Zstd, so the browser or crawler downloads fewer bytes. It's not a ranking factor, but it speeds up page loads and helps pages stay under crawler fetch limits. themselves live in the Image FormatsThe file encoding used to store an image (JPEG, PNG, WebP, AVIF, SVG, GIF) — it drives file size and Core Web Vitals, but carries no direct ranking weight. and Image Optimization guides, not here).
• The big underused point — don’t bulk-rename indexedStoring a crawled page in the search index so it can appear in results. Crawled is not the same as indexed — Google selects what to keep, and indexing isn't guaranteed. images. Mueller: “minimal effect, maybe no visible effect at all” if alt text/context are already good, and a mass rename is “annoying in Image Search” for a stretch — Mueller’s own “couple of months” is a rough guess from a secondary recap, not a verified figure.
• Renaming is a URL migration → audit new-name quality, reachability, redirectsA redirect sends browsers and crawlers from a requested URL to a different one. An HTTP redirect specifically is a 3xx status code paired with a Location header; meta refresh and JavaScript redirects achieve a similar navigation without being a 3xx response themselves. Permanent redirects (301/308) are Google's signal the target should be canonical; temporary ones (302/303/307) aren't., references, and declared-vs-actual type together, not naming alone. 301s needed, 404 risk, CDN/browser cache churn. Illyes framed image-URL-migration signal loss as “a few weeks.”
• Rename opportunistically (page rebuilds, migrations), never as a standalone SEO project. Expect not to be able to measure the gain.
• File names complement alt text; they don’t replace it. Alt text is the stronger primary signal and required for accessibilityWeb accessibility means designing sites so people with disabilities can use them, per the W3C's WCAG guidelines. It overlaps with SEO in specific, checkable ways — alt text, heading structure, descriptive link text, captions, and page speed all serve both audiences — but Google has said accessibility itself is not a ranking factor, and most WCAG success criteria (keyboard focus order, ARIA live regions, form labels) have no SEO effect at all.; screen readers fall back to the raw file name only when alt text is missing.

Official documentation

Primary-source documentation from the search engines.

Google

• Google Images best practices — the core reference, including the descriptive-filename rule, the my-new-black-kitten.jpg example, avoiding generic names, the alt-text keyword-stuffing warning, and localized-filename URL encoding.
• Let’s Talk Image SEO (Search Off the Record, Google, Oct 2022) — Lizzi Sassman and John Mueller on image naming, alt textAlt text is the value of the `alt` attribute on an HTML `<img>` element — a short text substitute chosen for the image's purpose and context, not a literal description of what it shows. It makes images accessible to screen-reader users and helps search engines understand images, mainly for image search., crawl frequencyCrawl frequency is how often a search engine comes back to re-fetch a page it already knows about. Popular pages that change often get refreshed many times a day; stable pages can go weeks or months between crawls — and you influence it indirectly, not by setting a dial., and why bulk-renaming indexedStoring a crawled page in the search index so it can appear in results. Crawled is not the same as indexed — Google selects what to keep, and indexing isn't guaranteed. images has minimal upside.

Bing / Microsoft

• Bing has no current, specific public guidance on file-name formatting (hyphens/underscores, length) comparable to Google’s Image SEOImage SEO is optimizing the images on your pages so search engines can discover, crawl, index, and rank them — in Google Images and visual search, and as part of standard web results. It spans file format, filenames, alt text, compression, responsive markup, structured data, and image sitemaps. page. Bing relies on file names and alt textAlt text is the value of the `alt` attribute on an HTML `<img>` element — a short text substitute chosen for the image's purpose and context, not a literal description of what it shows. It makes images accessible to screen-reader users and helps search engines understand images, mainly for image search. to understand images per its general image guidance, but there’s no usable verbatim Bing quote on this narrow point — a genuine gap rather than an omission.

Quotes from the source

On-the-record statements from Google. Where the source page exposes the text, the link is a deep link that jumps to the quoted passage.

Google docs — the file-name rule

• “The filename can give Google very light clues about the subject matter of the image.” Jump to quote
• “When possible, use filenames that are short, but descriptive. For example, my-new-black-kitten.jpg is better than IMG00023.JPG.” Jump to quote
• “Avoid using generic filenames like image1.jpg, pic.gif, 1.jpg when possible.” Jump to quote
• “If you localize your images, remember to also translate the filenames, keeping in mind the URL encoding guidelines.” Jump to quote

Google docs — the keyword-stuffing warning (scoped to alt textAlt text is the value of the `alt` attribute on an HTML `<img>` element — a short text substitute chosen for the image's purpose and context, not a literal description of what it shows. It makes images accessible to screen-reader users and helps search engines understand images, mainly for image search.)

• “Avoid filling alt attributes with keywords (also known as keyword stuffing) as it results in a negative user experience.” Jump to quote

John Mueller & Lizzi Sassman, Google — don’t bulk-rename indexedStoring a crawled page in the search index so it can appear in results. Crawled is not the same as indexed — Google selects what to keep, and indexing isn't guaranteed. images (Search Off the Record, via Search Engine Journal’s recap)

• Descriptive file names are worth doing, but the marginal upside is small: renaming “probably would have minimal effect, maybe no visible effect at all” if alt textAlt text is the value of the `alt` attribute on an HTML `<img>` element — a short text substitute chosen for the image's purpose and context, not a literal description of what it shows. It makes images accessible to screen-reader users and helps search engines understand images, mainly for image search. and surrounding context are already in place.
• On re-crawl slowness and the cost of a mass rename: images are crawled less often “because usually, they don’t change a lot,” so if “you changed all of them at once… over a period of a couple of months at least, it’ll be kind of annoying in Image Search.” SEJ recap

Gary Illyes, Google — image URL migration signal loss

• Signal loss from image URL migrations is “in line with web search results, which is a few weeks.” Coverage

Should I rename this image file right now?

The article’s guidance splits into three genuinely different paths depending on whether the image is new, already indexedStoring a crawled page in the search index so it can appear in results. Crawled is not the same as indexed — Google selects what to keep, and indexing isn't guaranteed., or about to be touched for an unrelated reason. Work through the question below to land on the right one.

Common mistakes to avoid

Seven concrete mistakes people make when naming image files, each with why it’s wrong and what to do instead.

• Bulk-renaming an existing, indexedStoring a crawled page in the search index so it can appear in results. Crawled is not the same as indexed — Google selects what to keep, and indexing isn't guaranteed. image library for a marginal SEO gain. Why it’s wrong: Mueller called the upside “minimal effect, maybe no visible effect at all” if alt textAlt text is the value of the `alt` attribute on an HTML `<img>` element — a short text substitute chosen for the image's purpose and context, not a literal description of what it shows. It makes images accessible to screen-reader users and helps search engines understand images, mainly for image search. and context are already good, and a mass rename causes weeks to months of Image Search disruption. What to do instead: leave indexed images alone unless you’re already touching them for another reason.
• Renaming without adding a 301 redirectA 301 redirect is the HTTP status code for a permanent move: it tells browsers and search engines a URL has moved for good, and it's the strongest signal for consolidating a page's ranking signals onto the new URL. Google says permanent redirects don't cause a loss in PageRank., leaving the old URL to 404. Why it’s wrong: the file name is part of the URL, so renaming breaks the old address for anyone or anything still pointing at it. What to do instead: set up a 301 from the old path to the new one every time you rename a live image.
• Using underscores instead of hyphens — red_running_shoes.jpg. Why it’s wrong: Google reads a hyphen as a word boundary but an underscore as joining words into one tokenA token is the smallest unit of text (or image/audio/video) an LLM processes — roughly 4 characters, or about ¾ of an English word. A context window is the maximum number of tokens (input plus output) a model can hold at once, like its short-term memory., so underscores communicate less (you won’t be penalized, but you lose a free signal). What to do instead: always separate words with hyphens.
• Generic names like image1.jpg, pic.gif, 1.jpg, or DSC_0491.jpg. Why it’s wrong: Google’s own guidance says to avoid generic filenames “when possible” because they tell Google (and anyone reading the URL) nothing about the subject. What to do instead: describe what’s actually in the picture, e.g. mount-rainier-sunrise.png.
• Keyword-stuffed names like shoes-buy-shoes-cheap-shoes-red-shoes.jpg. Why it’s wrong: it reads as spam and doesn’t help — one honest description is the goal. What to do instead: use a few real words that describe the image and stop.
• Treating the file name as a replacement for alt textAlt text is the value of the `alt` attribute on an HTML `<img>` element — a short text substitute chosen for the image's purpose and context, not a literal description of what it shows. It makes images accessible to screen-reader users and helps search engines understand images, mainly for image search.. Why it’s wrong: the file name is the lighter of the two signals, and alt text is required for accessibilityWeb accessibility means designing sites so people with disabilities can use them, per the W3C's WCAG guidelines. It overlaps with SEO in specific, checkable ways — alt text, heading structure, descriptive link text, captions, and page speed all serve both audiences — but Google has said accessibility itself is not a ranking factor, and most WCAG success criteria (keyboard focus order, ARIA live regions, form labels) have no SEO effect at all. — a screen reader only falls back to reading the raw file name when alt text is missing or broken. What to do instead: write proper alt text and give the file a descriptive name; don’t let one stand in for the other.
• Forgetting to translate or encode file names on localized pages. Why it’s wrong: Google explicitly says to translate filenames on localized images and to keep URL encoding guidelines in mind, since the file name is part of the URL. What to do instead: translate the descriptive words in the file name for each localized version and URL-encode non-Latin or accented characters.

Ready-to-copy prompts for naming image files

Paste an image’s context into an AI assistant to get several hyphenated, non-stuffed file name candidates that follow this article’s naming rules.

Prompt: generate file name candidates from an image’s context

Paste in what the image shows, its planned alt textAlt text is the value of the `alt` attribute on an HTML `<img>` element — a short text substitute chosen for the image's purpose and context, not a literal description of what it shows. It makes images accessible to screen-reader users and helps search engines understand images, mainly for image search., and the surrounding page topic. Expect back three to five short, hyphenated, lowercase file name options with no keyword stuffing.

I need a file name for an image on my website. Here's the context:

- What the image shows: [describe the subject, e.g. "a golden retriever puppy
  sitting on a wood floor"]
- Planned alt text: [paste the alt text you're using, e.g. "Golden retriever
  puppy sitting on a hardwood floor"]
- Page topic / surrounding text: [one sentence on what the page is about]

Give me 3-5 candidate file names that:
- Use real words separated by hyphens, not underscores
- Are lowercase and roughly 3-8 words long
- Describe the subject without stuffing keywords
- Don't duplicate the exact alt text word-for-word, but stay consistent with it

Format each as a ready-to-use filename with the .jpg extension.

Prompt: check a batch of existing file names for naming-rule violations

Paste in a list of current image file names. Expect back a flagged list noting which ones use underscores, are generic, or look keyword-stuffed — without any recommendation to rename images that are already indexedStoring a crawled page in the search index so it can appear in results. Crawled is not the same as indexed — Google selects what to keep, and indexing isn't guaranteed. and live.

Here is a list of image file names currently on my site:

[paste file names, one per line]

For each one, tell me if it:
- Uses underscores instead of hyphens
- Is generic (like image1.jpg, pic.gif, DSC_0491.jpg)
- Looks keyword-stuffed

Don't recommend renaming any of these if they're already live and indexed —
just flag which naming-rule issues each one has, so I know what to do
differently on new images going forward.

Test yourself: naming images for SEO

Five quick questions on how to name image files. Pick an answer for each, then check.

Changelog