Google updates Ads API to v24.1, expands mobile and AI testing
Serge Bulaev
Google has updated its Ads API to version 24.1, which may make it easier for advertisers to track mobile results, control ad creatives, and automate experiments. The update, released in May 2026, adds new tools for testing AI and video campaigns, and lets advertisers upload static images exactly as designed. Reporting for experiments appears to be more transparent with new data fields for confidence intervals and p-values. There are also new security features and a reminder that data older than 37 months will not be available after June 2026. These changes might help advertisers analyze results better and keep their data workflows up to date.

The latest Google Ads API v24.1 update, released in May 2026, gives advertisers powerful new tools to refine mobile campaign tracking, automate AI-driven experiments, and gain stricter control over ad creatives. This version centers on delivering deeper performance visibility and simplifying workflows ahead of the v22 sunset.
Deeper Mobile Reporting
The Google Ads API v24.1 update introduces advanced tools for advertisers, including granular mobile performance tracking by operating system, expanded A/B testing for AI and video campaigns, and enhanced creative controls. The release also adds new security features and enforces an upcoming data retention policy.
This version introduces a new segments.mobile_device_platform field, allowing advertisers to differentiate performance data between iOS and Android. This granular breakdown is critical for app-install and e-commerce campaigns, where conversion rates can vary significantly between platforms, enabling more precise budget allocation, platform-specific creative tailoring, and refined bidding strategies.
Expanded Experiment Support
The API now supports four new ExperimentTypeEnum values: ADOPT_AI_MAX, ADOPT_BROAD_MATCH_KEYWORDS, OPTIMIZE_ASSETS, and PMAX_REPLACEMENT_SHOPPING. These additions enable direct, side-by-side testing for AI Max, Demand Gen, and Performance Max campaigns. Reporting is also more transparent, with new fields like conversions_absolute_change_* that return statistical analysis for conversion metrics directly, eliminating the need for separate statistical analysis.
Creative Control in Demand Gen
Responding to advertiser feedback, v24.1 adds a classic_display_images field to DemandGenMultiAssetAd. This feature allows marketers to upload static images that will be used exactly as designed, preventing automatic asset recombination and ensuring strict adherence to brand guidelines.
Security and Compliance Additions
A new passkey_enabled boolean field now indicates if an account has adopted Google's passwordless authentication, helping developers encourage clients to upgrade their security. The update also serves as a reminder of the upcoming data retention policy changes that will affect queries for older granular data.
Key Field Requirements for Video Ads
To ensure ad integrity, several objects now have stricter requirements:
- DemandGenVideoResponsiveAdInfo: requires videos and logo_images.
- VideoResponsiveAdInfo: requires videos, business_name, and logo_images.
On a positive note, VideoResponsiveAdInfo is now mutable, allowing advertisers to update creatives without having to recreate the entire ad, which simplifies campaign management.
Developer Migration Checklist
- Upgrade all client libraries to v24.1 and validate new OAuth scopes.
- Rewrite
campaign_budgetqueries that use the removedad_sub_network_typesegment (ALM Corp guide). - Implement error handling for
DateRangeError.INVALID_DATEto prevent dashboard failures when data retention limits are enforced. - Integrate the new experiment metrics into reporting dashboards to leverage enhanced A/B testing capabilities.
By adopting these updates, advertisers and developers can unlock more precise OS-level insights, conduct statistically sound experiments, and ensure their data workflows are compliant with upcoming platform changes.
What performance insights become available at the mobile-device level in Google Ads API v24.1?
API v24.1 adds a new segments.mobile_device_platform field that separates performance by iOS vs. Android traffic for every campaign, ad group, and customer record. Industry reports suggest that conversion rates can differ significantly between the two operating systems, especially for app-install and e-commerce campaigns. This granular split lets advertisers:
- Shift budgets in real time when one OS starts outperforming the other.
- Tailor creatives (e.g., highlight Apple Pay buttons on iOS and Google Pay on Android).
- Refine bidding strategies by adjusting target CPA or ROAS per platform.
Which new experiment types are now supported by the API?
The release adds four new ExperimentTypeEnum values that run inside Google Ads accounts:
1. ADOPT_AI_MAX - tests Google's AI Max engine against standard broad-match keyword logic.
2. ADOPT_BROAD_MATCH_KEYWORDS - isolates the impact of switching to broad match only.
3. OPTIMIZE_ASSETS - rotates asset groups to find top-performing creatives.
4. PMAX_REPLACEMENT_SHOPPING - runs head-to-head Performance Max vs. traditional Shopping campaigns.
Each experiment now includes transparent reporting across clicks, conversions, impressions, and revenue, with built-in statistical analysis fields to speed up validation.
Do I need to update my existing integrations?
Yes. Although v24.1 introduces minimal breaking changes, two widely used objects became stricter:
- DemandGenVideoResponsiveAdInfo now requires both videos and logo_images.
- VideoResponsiveAdInfo is mutable for the first time, but it also requires business_name, videos, and logo_images.
If your queries still rely on the deprecated ad_sub_network_type segment tied to campaign budgets, those reports will fail starting with the upgrade; rewrite them to use the new segments.device or segments.conversion_action_name alternatives.
How does Google Ads API prepare developers for the upcoming data retention limits?
The API will implement data retention policies that will affect requests for older granular data, returning DateRangeError.INVALID_DATE errors instead of empty rows. The API documentation now flags this explicitly so you can build automated fall-backs. Google recommends:
- Exporting historical granular data into your own warehouse before it ages out.
- Switching to monthly segmentation for long-range trend analysis - monthly data is retained for longer periods.
- Backfilling BigQuery transfers early; backfills for older dates will be affected by the retention policy.
Where can I find the updated client libraries and code samples?
All official client libraries (Java, .NET, PHP, Python, Ruby, and Node.js) have been refreshed for v24.1. These can be found through Google's developer resources, which also include examples for running the new experiment types and mobile-device segmentation queries.