Previous value: -"With location or uule outside the US: always crawl from an IP in the request's own country. Without it, a batch page may be crawled from a US IP while country, language, location and the Google domain stay yours; exit_country in the result says which was used."New value: +"Internal, not documented: crawl from an exit in the request's own country."

Added value: +{
+  "description": "With location or uule outside the US: always crawl from an IP in the request's own country. Without it, a batch page may be crawled from a US IP while country, language, location and the Google domain stay yours; exit_country in the result says which was used.",
+  "type": "boolean"
+}

Previous value: -"With output \"html\": rewrite Google's encrypted /goto result links in the page to their real destinations, result titles first (adds up to about 10 seconds). Google limits how fast one page's links can be resolved, so about 5-6 links per page are resolved; the others keep Google's href. In parsed JSON results a link is never a Google redirect: one that could not be resolved is the site's home page and carries link_unresolved: true."New value: +"Resolve Google's encrypted /goto result links to their real destinations. Google lets a page's links be resolved all at once only a couple of minutes after serving it, so on async jobs and batches the job stays \"processing\" (links_status \"pending\") until all are resolved, about 3 minutes after the crawl, and then completes with one webhook. On the live queue it answers at once with about 5-6 links per page resolved, result titles first (up to about 10 seconds). With output \"html\" the page's hrefs are rewritten (unresolved ones keep Google's href). In parsed JSON a link is never a Google redirect: one not resolved is the site's home page and carries link_unresolved: true."

Previous value: -"With output \"html\": rewrite Google's encrypted /goto result links in the page to their real destinations (adds about 5-7 seconds). A link that cannot be resolved keeps Google's href. Parsed JSON results always have resolved links."New value: +"With output \"html\": rewrite Google's encrypted /goto result links in the page to their real destinations, result titles first (adds up to about 10 seconds). Google limits how fast one page's links can be resolved, so about 5-6 links per page are resolved; the others keep Google's href. In parsed JSON results a link is never a Google redirect: one that could not be resolved is the site's home page and carries link_unresolved: true."

Previous value: -"Advanced: a raw Google uule value (\"w+...\", \"a+...\", \"j+...\"), sent as is. Use location instead unless you already have uule values."New value: +"Advanced: a raw Google uule value (\"w+...\", \"a+...\", \"j+...\"), plain or URL-encoded as copied from a Google URL. Use location instead unless you already have uule values."

Removed value: -"^[A-Za-z][+ ][A-Za-z0-9+/_=-]{4,1000}$"

Added value: +"\"desktop\" (default), \"mobile\" (Android phone, Pixel 8), \"iphone\" (iPhone 15, Safari), \"ipad\" (iPad Air, Safari) or \"galaxy\" (Samsung Galaxy S24, Chrome)."

Previous value: -[
-  "desktop",
-  "mobile"
-]New value: +[
+  "desktop",
+  "mobile",
+  "iphone",
+  "ipad",
+  "galaxy"
+]

Added value: +{
+  "description": "Fetch pages 1..pages (up to 3) in this one request: the result has a \"pages\" array, one entry per page with its own status, credits and output (html_url, or the parsed results). 1 credit per page; a failed page, or a page whose AI Overview timed out, is free. wait_for_ai_overview applies to page 1. Not with page.",
+  "maximum": 3,
+  "minimum": 1,
+  "type": "integer"
+}

Previous value: -"Advanced: a raw Google uule value, sent as is. Use location instead unless you already have uule values."New value: +"Advanced: a raw Google uule value (\"w+...\", \"a+...\", \"j+...\"), sent as is. Use location instead unless you already have uule values."

Previous value: -"^[aw][+ ][A-Za-z0-9+/_=-]{4,1000}$"New value: +"^[A-Za-z][+ ][A-Za-z0-9+/_=-]{4,1000}$"

Previous value: -"Wait (up to about 10 seconds) for the AI Overview to finish loading before the page is captured."New value: +"Wait (up to about 10 seconds) for the AI Overview to finish loading before the page is captured. A page whose AI Overview does not load in time (ai_overview_status \"timeout\") costs 0 credits."

Added value: +{
+  "description": "With output \"html\": rewrite Google's encrypted /goto result links in the page to their real destinations (adds about 5-7 seconds). A link that cannot be resolved keeps Google's href. Parsed JSON results always have resolved links.",
+  "type": "boolean"
+}

Added value: +{
+  "description": "Wait (up to about 10 seconds) for the AI Overview to finish loading before the page is captured.",
+  "type": "boolean"
+}