There is no guaranteed timeframe, and Google does not offer one. Its own documentation says "it can take a few weeks for Google to notice a new site, or any changes in your existing site", and adds that "although Google crawls billions of pages, it's inevitable that some sites will be missed" (developers.google.com, checked August 2026). Anyone quoting you a confident "four to fourteen days" is repeating a number nobody measured on your site.
Which is the point worth making before any range: this is a question you can only answer for yourself, by measuring it. We added newest-first ordering to our Bulk Google Index Checker for exactly that reason, because the pages where indexing time actually matters are the ones you published this week, and Search Console will not tell you when they crossed over.
How long does Google take to index a new page?
For an established site that Google already crawls often, hours to a few days is typical. For a new domain with few links, weeks is normal and Google says so directly. For any given page, the honest answer is that it may never be indexed at all, because crawling and indexing are separate decisions and Google states plainly that "not every page that is crawled will necessarily be indexed".
Those three cases are not variations of one number. They are different situations with different bottlenecks, and the reason published estimates vary so wildly is that each writer measured a different one.
What reliably shortens the wait is not a trick. It is being a site Google already visits frequently, publishing something that merits a slot, and linking the new page from pages that get crawled often. Everything else is at the margins.
Why is there no single number?
Because "indexed" is not one event. It is three gates in sequence, each with its own delay and its own way of stalling, and a page can sit at any of them for reasons that have nothing to do with the next.
Conflating those gates is what makes the question unanswerable in the abstract. A page waiting at discovery needs a link or a sitemap entry. A page waiting at the crawl gate usually needs a faster server, not better writing, which is what "Discovered, currently not indexed" is really telling you. A page stuck at the index gate needs to be worth indexing.
Two forces set the pace across all three: how badly Google wants your pages, and how much load it thinks your server can take. Google calls the second one a crawl capacity limit, and documents that response times, 5xx errors and HTTP 429 responses all push it down.
What actually makes indexing faster?
Internal links from pages Google already crawls frequently are the cheapest lever, and usually the most effective. A new post linked from your homepage and from two established articles gets found on the next pass of pages Google visits anyway. A new post linked from nothing waits for a sitemap crawl.
An accurate sitemap lastmod is the second. It is a signal Google uses to decide what is worth revisiting, and it is only useful if it is honest: a CMS that stamps every URL with today's date teaches Google to ignore the field entirely, which costs you the one direct hint you had.
Server responsiveness is the third and the one nobody checks. If Google's crawl capacity limit for your site is low because responses are slow, everything queues behind that, including pages you just published. Search Console's Crawl Stats report shows what Googlebot actually measured, which is more trustworthy than a speed test run from a datacentre.
What does not work, despite being sold constantly, is pushing ordinary pages through Google's Indexing API. That API is documented for job postings and livestream events, and using it on regular content is outside its terms. We went through what each submission route can legitimately reach in instant indexing: IndexNow, Google's Indexing API, and what actually works.
Does "Request indexing" actually speed it up?
It helps at the discovery gate and does nothing at the others. Request Indexing puts a URL into a priority crawl queue, so if Google simply has not reached the page yet, it can bring the fetch forward. That is a real and useful effect.
It cannot make Google decide your page deserves indexing. If the page has already been crawled and declined, resubmitting asks for a re-evaluation of content that has not changed, and Google's guidance for that status ends by telling you there is no need to resubmit at all.
The practical rule: use it once for a genuinely new or genuinely changed page, and never as a repeated ritual. Submitting the same unchanged URL five times produces exactly what submitting it once produced.
How do I measure this on my own site?
Record the publish time, then check indexing status on a schedule until it flips. That is the whole method, and it is the only way to get a number that describes your site rather than someone else's.
The reason almost nobody does it is that Search Console makes it awkward. The Page Indexing report lags by days, so by the time a page shows as indexed you have lost the moment it happened, and its example URL lists are capped at 1,000 URLs. You cannot reconstruct a timeline from a report that reports late and shows you a sample.
Checking directly gets you the timestamp. The URL Inspection API returns a live verdict per URL within a quota of 2,000 inspections per property per day, which is far more than enough to poll a handful of recent posts daily. Run it once a day against your newest pages and after a month you have your own median time to index, measured rather than borrowed. The quota arithmetic behind that, including what happens on larger sites, is in the URL Inspection API's 2,000-a-day limit.
When should I stop waiting and start diagnosing?
Two to four weeks for a single page on an established site. Before that, waiting is the correct action and intervening usually means rewriting something that was going to be fine.
Volume changes the reading completely. One page waiting is noise. Twenty pages from the same template waiting is a pattern, and the pattern is the diagnosis: read the status rather than the delay, because "Discovered" and "Crawled" point at opposite fixes and neither is about patience.
The number worth watching is not how long any single page took. It is how many of the pages you published this month are still missing, which is a work list rather than a statistic, and which is why the total count of indexed pages is close to meaningless on its own.
FAQ
Does Google guarantee it will index my page?
No. Google's documentation states that although it crawls billions of pages, "it's inevitable that some sites will be missed", and separately that not every page which is crawled will necessarily be indexed. Crawling and indexing are two different decisions: being fetched does not entitle a page to a place in the index. There is no submission method, plugin or paid service that changes this, because the decision is editorial rather than technical.
Why did my new page get indexed in hours but another took weeks?
Because they almost certainly stalled at different gates. A page linked from your homepage is found on the next crawl of a page Google already visits often, while a page linked from nothing waits for a sitemap crawl. Beyond discovery, Google's crawl capacity limit for your site affects how quickly it fetches anything, and the index decision itself is made per page on quality and duplication grounds. Same site, same day, entirely different paths.
Does submitting a sitemap make Google index pages faster?
It helps discovery, which is only the first of three gates. A sitemap tells Google the URL exists and, through an accurate lastmod date, hints that it changed. It does not make Google fetch sooner if crawl capacity is the constraint, and it has no influence on whether the page is judged worth indexing. An inaccurate lastmod is worse than none: if every URL claims today's date, Google learns to ignore the field.
How can I tell how long my own site takes to index pages?
Record the publish time and check indexing status daily until it changes, using the URL Inspection API rather than the Page Indexing report. The report lags by days and caps its example URL lists at 1,000, so by the time it shows a page as indexed the actual moment is lost. Polling a handful of recent posts each day costs a tiny fraction of the 2,000 daily inspections Google allows per property, and after a month you have a median measured on your own site instead of a figure borrowed from an article.