SEO Tools
Hreflang Generator & Validator
Generate alternate-language tags and check supplied annotations for invalid codes, duplicates, self-references, and return links. All processing stays in your browser.
Build your language group
Add equivalent versions of one page. Checks and generated tags update automatically.
Up to 100 rows. Examples: en, en-GB, id, zh-Hant, zh-Hans-US, x-default. Use GB rather than UK. URLs may be on different domains. A fallback may share an existing language URL.
Import language rows from CSV
Required headers: hreflang,url. Import replaces these rows after CSV parsing succeeds. Invalid codes or URLs remain editable and block generation.
Generated HTML tags
Place the complete identical set inside the head of every listed page, including each page’s self-reference. Generated tags describe your inputs; page content, installation, and return links are not verified.
Inspect supplied annotations
Paste HTML link tags or import page annotations from CSV. Nothing is crawled.
Limits: 512 KiB UTF-8, 2,000 annotation/CSV records, 200 source pages, 10,000 HTML elements. CSV headers: source_url,hreflang,target_url. A source-only row with blank hreflang and target_url declares a supplied page with zero annotations. CSV record numbers include the header; HTML record numbers count extracted link tags.
Variants
Issues
Review notes
Output
Checks across supplied pages
Missing return links are reported only when the target page is supplied. A missing target page means not checked. CSV records must contain the complete annotation sets for those pages; omitted records can appear as missing links.
| Record | hreflang | Source page | Alternate URL |
|---|
Findings and guidance
Use hreflang with confidence
Which language codes are supported?
Use a two-letter ISO 639-1 language, an optional named ISO 15924 script, and an optional ISO 3166-1 alpha-2 region, in that order. Examples: en, pt-BR, zh-Hant, sr-Latn-RS. x-default is a separate fallback value. This tool uses bundled lookup lists; numeric regions such as es-419, private extensions, reserved regions, and non-linguistic script markers are unsupported. It cannot detect the actual language of your content.
How do self-references and return links work?
Each page’s annotation set should point to itself and its equivalent alternate versions. If page A points to B, B should point back to A. Use the CSV format to supply complete sets for multiple pages. HTML mode inspects one supplied page; the other pages are not fetched. A valid URL is compared using browser URL normalization; path case, trailing slashes, query strings, and query order remain significant.
Is x-default required?
No. It is an optional fallback for unmatched language or region settings. It can point to a language selector or another appropriate page, including a URL already used by a language annotation. A generic language version can also help users outside your targeted regions when it matches your publishing setup.
What does this tool not verify?
This is a static check of supplied data. It does not check HTTP headers, sitemap annotations, live availability, redirects, canonical tags, indexability, actual translations, or correct deployment in a well-formed HTML head. Browsers may repair malformed HTML. Passing these checks does not guarantee search-engine acceptance, rankings, or AI citations.
Do inputs leave my device?
No. Local files are decoded as UTF-8, supplied HTML is parsed in an inert template, and generated output is displayed as text. There are no processing requests, remote lookups, uploads, or browser storage. Changes clear stale reports; invalid file reads keep existing inputs.
Reference: Google’s localized-page guidance.
