Firebase AI Logic

Firebase AI Logic supports the latest Gemini stable model: gemini-3.5-flash. You can directly access this model in your mobile or web app via either the Gemini Developer API or the Vertex AI Gemini API.

• If you're using the Gemini Developer API, accessing this model does not require the pay-as-you-go Blaze pricing plan. If you access the model via the Vertex AI Gemini API, make sure to set the location in your request to global.
• If you're migrating from the preview version of this model:
  • Update model name: gemini-3-flash-preview → gemini-3.5-flash
  • Review pricing. Gemini 3.5 Flash is more expensive than Gemini 3 Flash Preview. See pricing information: Gemini Developer API | Vertex AI Gemini API.
  • Test your prompts. Default thinking level changed from HIGH → MEDIUM; verify quality, speed, and cost.
  • Remove temperature, topP, topK from your config. Note that these parameters are no longer recommended for any of the Gemini 3.x and later models. For more deterministic responses, define system instructions with explicit rules for your specific use case.

Cloud Functions

Updated documentation regarding default deployment regions for Cloud Functions.

Firebase CLI 15.17.0 and higher now deploys functions based on your project configuration and the region of the triggering data source (such as Cloud Firestore or Cloud Storage) whenever possible, rather than unconditionally defaulting to us-central1. You are encouraged to explicitly set regions in your function definitions. This change does not affect your existing deployed functions.

SDK Releases

The Firebase SDK for Flutter (v4.13.0) is now available. This release includes changes in several products, and bug fixes in several products.

Firebase AI Logic

Firebase AI Logic now lets you specify an on-device model when using hybrid / on-device inference in Android apps. With this capability, you can access preview on-device models, like Gemini Nano 4 (which is based on Gemma 4).

SDK Releases

The Firebase Android SDKs for Firebase AI Logic, Authentication, Cloud Firestore, Cloud Messaging, Crashlytics, Crashlytics NDK, Firebase Phone Number Verification, Remote Config, and SQL Connect have been updated. For more details, check out the latest Android Release Notes. To get started with Firebase in Android, see Add Firebase to your Android Project.

The Firebase Android BoM (Bill of Materials) has been updated to v34.13.0. Learn how to add SDKs to your app using the Firebase BoM.

The Firebase JavaScript SDK (v12.13.0) is now available. For more details, see the Firebase JavaScript SDK release notes. To install the SDK, see Add Firebase to your JavaScript Project.

The Firebase Admin Node.js SDK (v13.9.0) is now available. This release deprecates Node.js 20 support and comes with a new feature in the Remote Config API. To install this SDK, see Add the Firebase Admin SDK to your server.

Firebase AI Logic

Firebase AI Logic server prompt templates now support multi-turn interactions (including chat, iterative editing of images with Gemini Image models, and function calling).

Server prompt templates allow you to store your prompt, schema, and configurations server-side. Then in the request from your app, you only provide the template's ID and any desired user-defined inputs. You can protect against exposing your prompt client-side, as well as easily update your prompt, schema, and configs without releasing a new app version.

Cloud Firestore

You can now use pipeline operations to perform joins with subqueries. To learn more, see Perform joins with subqueries.

Pipeline operations are now Generally Available (GA). To learn more, see Get started with pipelines.

Firestore with MongoDB compatibility now supports text and geospatial search. You can create text indexes, perform text and geospatial search queries using the $text and $near operators, handle language settings, and calculate relevance scores. To learn more, see Text search and Geospatial search.

Cloud Firestore Enterprise Native databases now support text search and geospatial search. To learn more, see Text search and Geospatial search.

Pipeline operations now support the delete and update output stages. To learn more, see the Delete and Update stage references.

The maximum document size for Firestore with MongoDB compatibility databases has been increased to 16 MiB. To learn more, see Behavior differences and Quotas and limits.

You can now perform bulk delete operations, such as dropping a collection, in Cloud Firestore Enterprise Native and Firestore with MongoDB compatibility databases. To learn more, see Bulk delete data and Bulk delete data in Firestore with MongoDB compatibility.

The Cloud Firestore Emulator now supports Enterprise feature development using the Firebase CLI. To learn more, see Connect your app to the Cloud Firestore Emulator.

Firestore with MongoDB compatibility now supports change streams. Change streams let applications access real-time changes (inserts, updates, and deletes) made to a collection or to an entire database. To learn more, see Change streams.