IndexNow is an open protocol that allows websites to notify search engines the instant new content is published — rather than waiting for the search engine's crawler to passively discover the new URL during a scheduled crawl. Supporting search engines include Microsoft Bing, Yandex, and Naver. Because Bing's index feeds Bing Copilot, Perplexity, and other AI retrieval systems, IndexNow submission is one of the most direct ways to accelerate a new article's appearance in AI answers.
How IndexNow Works
The IndexNow protocol requires two components: an API key file hosted on the domain, and an API call to the IndexNow endpoint when new content is published. The API key is a text file placed at the root of the domain (e.g., yourdomain.com/your-api-key.txt). The submission call sends the URL to be indexed to the IndexNow API endpoint along with the API key. The search engine validates that the API key file is hosted on the domain and then prioritizes crawling the submitted URL.
Automating IndexNow With GitHub Actions
The most effective IndexNow implementation for a site that publishes content through a Git-based workflow is a GitHub Action that runs on every push to the main branch. The action generates the updated sitemap, extracts the new URLs from the sitemap, and submits them to the IndexNow endpoint. This means every content deployment automatically triggers IndexNow submission — no manual step required, no delay between publication and search engine notification.
What IndexNow Does Not Do
IndexNow notifies search engines that a URL is ready to crawl. It does not guarantee immediate crawling or indexing — it prioritizes the URL in the crawl queue. For most sites, the prioritized crawl happens within hours. For Google specifically, IndexNow is not directly supported — Google uses its own URL submission methods, though Google's crawlers often pick up content quickly through Bing Copilot's index signals.
The Omni AEO infrastructure setup includes IndexNow auto-submission as a standard GitHub Action — implemented once and running automatically on every subsequent deployment.