Contentstack Adds AI Tagging, Duplicate Detection and Self-Service Tiers to Assets Add-On

Contentstack on Sept. 23 released a set of updates to its Assets add-on that give content teams more control over metadata, automate tagging with artificial intelligence and surface duplicate images for cleanup, according to a report by CMSWire published Sept. 28, 2026.

The Assets release covers five capabilities: a select-type user-defined field for consistent metadata entry; background duplicate-image detection that groups identical images for review; AI-generated tags and alt text applied at upload without overwriting user-supplied values; AI-suggested metadata exposed programmatically via API; and persistent user preferences that follow users across browsers and devices. Contentstack said the duplicate-management feature is rolling out gradually and may not yet be available to all customers. Assets remains a paid add-on.

Alongside the Assets updates, Contentstack introduced self-service account creation and subscription management. Users can now sign up and choose from Free, Build or Growth tiers and manage billing without contacting sales. The company described this as suited to smaller teams that want to start building and scale at their own pace.

Headless CMS Updates

Beyond Assets, Contentstack shipped several updates to its headless CMS in September. A Sept. 15 update lets editors select every entry a search or filter returns — not just those on the current page — and then publish, unpublish, delete or add to a release in a single pass. A single action covers up to 10,000 entries and runs as one background job trackable in the Stack Bulk Task Queue. The select-all function does not support bulk workflow updates or export.

On Sept. 11, Contentstack added support for up to five SAML 2.0 identity providers within a single organization, each with its own certificate, session policy, role mapping and login URL. The company indicated the feature addresses organizations whose users span separate identity systems — such as a parent company and an acquired subsidiary on different tenants, or employees and contractors on different providers. The feature is not enabled by default; organizations must request it through Contentstack support.

Also on Sept. 11, Contentstack changed how its Content Delivery API and Content Management API handle the limit query parameter, which now accepts positive integers only. Previously, passing limit=0 returned every matching record; it now returns the default number of records, which is 100 per request. According to the company, the old behavior could cause requests on large stacks to time out. Integrations that relied on limit=0 must now paginate using the skip parameter.

A Sept. 7 update lets developers resolve nested embedded entries and assets in the JSON Rich Text Editor in a single GET request using include_embedded_items[]=RECURSIVE, with depth capped at five levels. Contentstack said existing behavior is unchanged unless developers opt in.

Broader Context

Contentstack made what CMSWire described as its most consequential strategic move in June, launching its Agentic Experience Platform (AXP) — a deliberate break from the digital experience platform (DXP) category. The architecture is built on three layers: Content Cloud, Data Cloud and Agent OS. In early 2026, Contentstack earned recognition as a Visionary in the 2025 Gartner Magic Quadrant for Digital Experience Platforms and as a Leader in the Forrester Wave CMS report. The company is scheduled to bring the AXP narrative to customers at ContentCon Chicago 2026 on Oct. 27–28.