What Google suggests as you type
keyword_autocomplete is a $0.20 GET request on the Wuthering AI API. Read the live suggestions Google offers for a phrase, exactly as a searcher sees them. It answers complete JSON, and a call that fails is not charged.
Endpoint
GET /v1/keyword_autocomplete
Price per successful call
$0.20
Failed calls
Not charged
curl 'https://wutheringai.com/v1/keyword_autocomplete?keyword=…&token=YOUR_TOKEN'
What does keyword_autocomplete return, and when should you call it?
Return the suggestions Google's search box is offering for a phrase right now, in the order it offers them, with a relevance score on each.
These are queries people are completing today, served by the engine itself, which makes this the one discovery tool here that can see a term before an index does: a product launched last week, a model released yesterday, a phrase spiking this morning. Every other keyword tool reads a database built from past crawls, so a term none of them has heard of still has suggestions here.
It reports shape, never size. There is no search volume, no cpc and no difficulty in the answer, and rank_group is the order the box offered a suggestion, NOT how often it is searched: on one fifteen-suggestion list, the eighth suggestion turned out to be searched fourteen times more often than the seventh. Price what comes back with wuthering_keyword_overview before ranking any of it.
Returns: suggestion, relevance (500-2000, higher is closer to the phrase you sent), suggestion_type, rank_group, rank_absolute, search_query_url. The thumbnail_url and highlighted fields are part of the row and arrive null.
One phrase per call, about fifteen suggestions back, so a real expansion is many calls with shaped seeds rather than one call with a long list: complete 'espresso machine for', 'espresso machine vs', 'best espresso machine' and the seed followed by each letter, and budget a call apiece.
Two things a caller assumes and should not. A suggestion NEED NOT CONTAIN the phrase you sent: given a phrase it has few completions for, the box answers with loosely related or entirely unrelated queries rather than with nothing, so filter on your own term before treating the list as expansion. And a misspelling is silently corrected — 'projct managment softwre' comes back with the suggestions for the corrected phrase and nothing in the response says so, which quietly turns a typo into a different market.
Use it to learn how searchers actually phrase a thing, and to reach terms too new to be indexed. Don't use it to size demand, which is wuthering_keyword_volume, and don't use it to expand one term with volume already attached, which is wuthering_keyword_suggestions.
What parameters does keyword_autocomplete take?
keyword_autocomplete accepts 4 parameters, of which 1 is required. Array values are comma-separated on the query string.
| Parameter | Type | Required | Description |
|---|---|---|---|
| keyword | string | yes | The phrase to complete, written the way a searcher would have typed it so far. Completions are built from what follows it, so the shape you send decides which half of the market you see: 'espresso machine' returns buying modifiers, 'how to espresso machine' returns questions. |
| cursor_position | number | no | Where the caret sits inside the keyword, counted in characters. Defaults to the end, which completes the phrase forwards ('espresso machine' to 'espresso machine with grinder'). Set it to 0 to complete backwards instead and get the words searchers put BEFORE the phrase ('breville espresso machine', 'best espresso machine') — a different set of terms that no other tool here returns. |
| location_code | number | no | Numeric location code to scope results to. Defaults to 2840 (United States). Search volume and competition vary sharply by country, so set this when the question is about a specific market.Defaults to 2840 |
| language_code | string | no | ISO language code for results, for example 'en' or 'de'. Defaults to 'en'.Defaults to en |
Reference for keyword_autocomplete · last updated · published by Wuthering AI