Google Adds Branded and Multimodal Data to Search Console API
Google has added three Search Console Performance report capabilities to the Search Analytics API, giving developers and reporting-tool providers new ways to retrieve search performance data. The additions cover branded and non-branded query classification, multimodal search performance, and requests that include multiple countries or device types.
The changes bring API functionality closer to options already available in the Performance report. Google says the expanded API can reduce the need for separate requests or manually maintained keyword regex lists. The features are subject to specific data availability and request limitations, so teams will need to account for those when updating reporting workflows.
Branded classification has a defined start date
A new brandedClassification dimension allows API users to filter or group results as BRANDED or NON_BRANDED. The classification data starts on Feb. 21, 2026, and is available only for top-level properties. Properties that do not meet Google’s minimum query and impression requirements may return an empty table rather than rows.
The API reference also sets limits on how the dimension can be used. It accepts only the equals operator, and requests that apply a query filter cannot use branded classification for filtering or grouping. As a result, teams should check the supported request combinations before changing existing reports.
The start date matters for historical comparisons: the new API classification cannot provide a branded-versus-non-branded breakdown for dates before Feb. 21, 2026. Google has said it uses an AI-assisted process to classify searches, and has cautioned that some queries may occasionally be misidentified. Teams that already use regex rules to identify brand terms can compare those results with the API classification to spot differences, while recognizing that the two approaches may not match exactly.
Multimodal searches receive an API type
The API now includes webMultimodal as a search type. Google’s reference says this covers performance metrics for searches using Google Lens, Circle to Search on Android, and image uploads to Google Search. This gives API users a distinct way to request multimodal performance data rather than relying solely on the reporting interface.
The timing and scope of the historical data remain unclear. Google’s API announcement and reference do not specify how far back multimodal data extends. The API materials also do not mention Chrome’s “Search this image” feature, although Google’s Performance report help information includes Chrome image search. The difference is a reason for teams to verify which activities are represented before interpreting the API figures as a complete measure of image-based search activity.
Google has also said multimodal data will be available in BigQuery for properties with an active bulk data export, but did not provide a date. As of Nov. 16, the bulk export table reference listed six search types and did not yet include a multimodal value. Google’s API documentation likewise does not clarify whether the standard web results include any webMultimodal data. Teams should compare API totals with the corresponding Search Console report figures before adding web and multimodal results together.
Multiple countries or devices in one request
The new filterLists field lets users filter results by multiple countries or multiple devices in a single API request. Google’s examples include regional groupings such as EMEA, LATAM and DACH, which may simplify reporting for organizations that currently make separate requests for individual markets.
The field is restricted to the country and device dimensions and uses exact-value matches. A request cannot use the same dimension in both its regular filters and filterLists; for example, country cannot appear in both places in one request. These constraints are important for teams designing regional or device-level reporting templates.
What reporting teams should check
The additions make it possible to bring more Performance report distinctions into automated reporting, but they do not remove the need to validate data coverage. Before rolling out changes, API users should confirm that their property is eligible for branded classification, account for its Feb. 21, 2026 start date, and test how the new dimensions interact with existing filters and groupings.
For multimodal reporting, the open questions include the available historical range, BigQuery availability and the relationship between web and webMultimodal results. Checking the API reference and bulk export documentation for updates—and comparing API output with the Performance report—can help teams avoid gaps or double counting as they revise their measurement systems.
