Detect the query-correction state shown on one public Google Search page. The Bot supports corrected-results, did-you-mean, multiple-suggestion, and valid no-correction states while excluding search-result content.
What Does BrowserAct Google Spell Check API Do?
The Bot builds one Google Search URL internally from the query and selected website options, opens the public result page, identifies only Google's visible query-correction state, extracts its visible fields, and stops when the correction banner or stable no-correction state is evaluated.
Atomic Capability
One run opens one public Google Search results page and returns only Query correction.
The Bot does not combine adjacent search features, open destination websites, or turn the request into a broad crawl. This narrow boundary makes the result predictable for direct API use and composition with other Bots.
Key Features
- Extracts one narrowly defined public Google Search capability.
- Accepts an open-ended query with a working default.
- Uses dropdowns for country and interface language.
- Builds the fixed Google Search URL and website codes internally.
- Returns the same top-level response envelope for positive and valid negative results.
- Preserves additional visible target-module fields in
unclassified_fields.
Capability Availability
| State | Output behavior |
|---|---|
| Target module is present | available is true; data contains the extracted result. |
| Valid result page without the target module | available is false; data=null. |
| Additional fields appear inside the target module | Values are preserved in unclassified_fields. |
Quick Start
- Open Run task.
- Enter a query, such as
Coffeee. - Choose the available country and language options.
- Run the Bot and review its structured response.
- Use the API Tab when you want to call the published template from an application or automation.
Input Parameters
| Parameter | Configuration |
|---|---|
query |
String · Required Default: CoffeeeOpen-ended Google search text; 1-300 characters after trimming. |
search_country |
Dropdown · Required Default: United StatesSelect a supported country; the Bot maps the label internally. |
interface_language |
Dropdown · Required Default: EnglishSelect a supported interface language; the Bot maps the label internally. |
See the Inputs Tab for supported labels, mappings, constraints, and the default request JSON.
Output Structure
| Field | Type and behavior |
|---|---|
capability |
Fixed string google_spell_check. |
available |
Boolean indicating whether the target module appeared. |
request_metadata |
Normalized public inputs and the internally constructed source URL. |
data |
One structured target-module object, or null when absent. |
unclassified_fields |
Additional visible fields from the same module. |
See the Output Tab for the complete schema and the actual response from a current Live Version run.
Example Output
This table shows one result from validated task 116386329846777035. For repeating records, only the first record is expanded in this preview.
| Field | Result |
|---|---|
capability |
google_spell_check |
available |
true |
request_metadata.original_query |
Coffeee |
request_metadata.search_country |
United States |
request_metadata.interface_language |
English |
request_metadata.source_page_url |
https://www.google.com/search?q=Coffeee&gl=us&hl=en |
data.original_query |
Coffeee |
data.displayed_query |
Coffee |
data.correction_type |
corrected_results |
data.corrected_query |
Coffee |
data.alternative_queries |
[] |
data.showing_results_for |
Coffee |
data.search_instead_text |
Coffeee |
data.search_instead_url |
https://www.google.com/search?sca_esv=6b332504a115b4d9&gl=us&hl=en&q=Coffeee&nfpr=1&sa=X&ved=2ahUKEwi9j53vudSWAxVPRTABHdvwKaYQvgUoAXoECBQQAg |
unclassified_fields |
{} |
How Does It Work?
- BrowserAct validates the query and maps each selected website option.
- The Bot constructs and opens one public Google Search results URL.
- It identifies only Google's visible query-correction state and reads its visible fields.
- It normalizes the result into the fixed
google_spell_checkresponse envelope. - It returns one structured response and stops without opening destination websites.
Response Behavior and Limits
- Target absent:
available=falsewithdata=nullis a valid response. - Conditional fields: A field may be null when the page does not display it; this does not automatically require rebuilding the Bot.
- Public access: The Bot reads a public Google Search page and requires no account login.
- Execution boundary: One run reads one selected search page and one target module.
- Destination pages: Returned links are not opened.
Use Cases
- Add the target search capability to an application without designing a browser workflow from scratch.
- Feed structured search data into monitoring, research, enrichment, or reporting systems.
- Compose this Bot with other BrowserAct templates through API, n8n, Make, Zapier, or MCP.
How to Build a New API-Ready Bot with Agent Built
Step 1: Define One Atomic Capability
Describe one public-web action, its user inputs, allowed navigation, output contract, and exact stopping point. Keep fixed URLs internal and give every visible input a working default.
Step 2: Let BrowserAct Explore the Capability Variants
Build the Bot against the live website and check representative positive triggers, one valid negative trigger, and at least one non-default dropdown path.
Step 3: Validate the Stable Contract
Publish the validated version, run it with the default inputs, and confirm the response envelope, actual output, execution time, credits, and API call before listing the template.
Automation and Export
BrowserAct produces structured data first. Use the completed result through Run Task, the JSON API, n8n, Make, Zapier, MCP-based agents, or an available export flow.
Need Help?
Cannot find the exact atomic capability you need, or need help composing several Bots?
Discord: BrowserAct Community
Email: service@browseract.com
