Firestore at Next ’26: Unlock agentic development, search and MongoDB compatibility

In the era of AI agents, the distance between a big idea and a working application has never been shorter. As we lean more heavily on agents to help us build applications, a critical question remains: can your database infrastructure keep up?

With its virtually limitless scalability and high availability, Firestore, Google Cloud’s fully managed document database, is a great fit for emerging agentic applications. And at Google Cloud Next ‘26, we leveled up Firestore for AI-driven apps even further, with:

• Tighter agentic AI integrations: New native integrations with AI Studio and third-party coding agents mean your LLMs and database now speak the same language.

• Full-text search and expressive queries: Differentiated search capabilities and pipeline operations mean agents and users are able to find exactly what they need within your data.

• Enhanced MongoDB compatibility: Now it’s easier than ever to bring existing MongoDB workloads into the Firestore ecosystem.

In this blog, we’ll take a closer look at our announcements from Next ‘26. But first, here’s a Firestore refresher. 

The case for Firestore

Whether you’re an enterprise leader looking to empower your workforce to build their own productivity tools, or a founder sketching out the next big thing on a napkin, you need to be able to prototype at the speed of thought, pivot the moment you get user feedback, and do it all without breaking the bank — or the database.

When it comes to selecting a database, you need to worry about:

• Scaling: Can the database survive a viral traffic spike?

• Budget efficiency: Does the solution scale to zero during inactivity to reduce your costs?

• Iteration speed: Will frequent tweaks in your agent prompts be slowed by expensive database schema migrations to fulfill those requests?

We designed Firestore to address these exact concerns. Firestore has always been an easy way to achieve rapid, automatic, elastic database scaling, with its serverless architecture that also provides sub-second provisioning. Meanwhile, Firestore’s document model makes it easy and fast to iterate on your data structures — no breaking schema changes, no downtime, just flow.

At the same time, accelerating development velocity shouldn’t mean compromising on enterprise governance. Firestore offers an industry-leading 99.999% SLA and ACID-compliant transactions, all while benefiting from the rigorous security and privacy oversight, fundamentally inherent to Google Cloud.

Companies like FlutterFlow are already reaping the benefits. 

“As an AI-native company dedicated to democratizing web and mobile development, Firestore has served as the foundational database powering FlutterFlow as we scaled from zero to over 3 million users across more than 150 countries. Over the past five years, we have experienced zero outages while serving more than 750 billion reads and 75 billion writes. We are true believers in Firestore.” – Abel Mengistu, CEO and Co-founder, FlutterFlow

With that background, here’s what’s new in Firestore from Next ‘26.

1. Accelerating application development through agentic AI integrations

We embedded Firestore directly into the AI creative process. Through new native integrations with AI Studio, developers can now build and provision fully functional full-stack applications with an integrated Firestore database and added authentication from a single natural language prompt. This integration is driving incredible momentum on Firestore, bringing the overall Firestore developer base to 750,000 monthly active developers and over 10M hosted databases.

We also enhanced the ability to integrate Firestore with preferred third-party AI agents, including Claude Code, Cursor, and Codex. With the general availability of Firestore Skills and Firestore remote MCP service, connecting to popular external agents is even more straightforward.

To further enhance productivity, we introduced natural language querying in the Google Cloud console, in preview. This leverages Gemini Code Assist to convert simple natural language queries into complex, MongoDB-compatible queries.

2. Differentiated search and queries

Building sophisticated, data-rich AI agents requires a database with modern search and query capabilities. Our reimagined query engine on Enterprise edition, featuring pipeline operations, is now generally available, and delivers hundreds of new query capabilities, positioning Firestore as a premier service for expressive applications. 

A major addition is built-in full-text search, now available in preview. Firestore full-text search leverages Google search technology, ensuring users who perform a search receive precise results using high-quality relevance models that support more than 40 languages. Moreover, alternative hybrid database and search setups can produce search results that aren’t reflective of actual database data, because their search indexes only use eventual consistency. In contrast, Firestore search indexes are strongly consistent with transactional data, for more accurate search results. Crucially, this native functionality inherits Firestore’s serverless architecture, drastically reducing the operational friction of managing separate search infrastructure.

We also introduced geospatial queries capabilities in preview, enabling developers to build location-aware applications that can easily find nearby points of interest.

This release also includes the highly requested JOIN functionality, in general availability. Implemented via subqueries, pipeline operations enable lookups across diverse collections. Additionally, we launched a preview of built-in data manipulation operators to facilitate the bulk normalization, sanitization, and backfilling of documents within your collections.

We’re also provided deeper observability insights through enhanced usage monitoring, including usage by collection through a new Usage Insights feature in preview.

Finally, Firestore will soon be integrated with Knowledge Catalog, providing you with deeper insights into how your data models evolve at the collection level.

3. Enhanced MongoDB compatibility and scalability

We continue to broaden Firestore’s appeal for enterprise workloads with enhanced MongoDB compatibility, making it easier for you to migrate and build on Firestore.  

To boost MongoDB compatibility, Firestore now supports larger documents up to 16MiB, removing traditional barriers for complex data migrations and high-volume workloads. 

To enable real-time data movement, we launched highly scalable change streams to synchronize changes from Firestore to services like BigQuery at scale. This feature is built to handle virtually any volume of read and write operations, giving you the piece of mind that change streams will seamlessly scale alongside database production traffic.

We also improved data lifecycle management, giving developers the ability to efficiently manage data deletion by dropping a collection and using more flexible time-to-live (TTL) time offsets for automatic data expiration — all while ensuring these administrative operations never impact the database’s production traffic.

Get started on Firestore

These new capabilities are now available with the Firestore Enterprise edition, available in both Native and MongoDB compatibility modes. Developers can begin incorporating these advanced capabilities into your intelligent, agentic applications today.