Firebase Updates & Release Notes
199 updates curated from 1 source by the Releasebot Team. Last updated: Nov 19, 2026
- Nov 18, 2026
- Date parsed from source:Nov 18, 2026
- First seen by Releasebot:Nov 19, 2026
November 18, 2026
Firebase AI Logic now supports the stable Gemini 4 Neon model in mobile and web apps through either Gemini API provider, with support for Office and OpenDocument file inputs. The model requires Blaze billing and has specific limits for thinking, caching, and request parameters.
Firebase AI Logic Firebase AI Logic supports the latest Gemini stable model gemini-4-neon. You can directly access this model in your mobile or web app via either the Gemini Developer API or the Agent Platform Gemini API (formerly Vertex AI). • Billing: Accessing gemini-4-neon requires the pay-as-you-go Blaze pricing plan regardless of your Gemini API provider. • Short-term availability: For the Agent Platform Gemini API (formerly Vertex AI), gemini-4-neon is a short-term availability model. To avoid service disruptions when models are updated, we strongly recommend using Firebase Remote Config or server prompt templates to dynamically manage the model name in your app. • If you're migrating from previous models, note the following for this new model: • Thinking levels (HIGH default, no XHIGH or thinkingBudget): • Supported thinking levels in Firebase AI Logic are LOW, MEDIUM, and HIGH (default). Note that Firebase AI Logic does not currently support XHIGH. • Does not support setting thinking level of MINIMAL (the request will fail with a 400 error). • Does not support using thinkingBudget (there's no backwards-compatibility; the request will fail with a 400 error). • Context caching: • Supports implicit caching by default (with a minimum of 4,096 input tokens on the Gemini Developer API and 6,144 input tokens on the Agent Platform Gemini API (formerly Vertex AI)). • Does not support explicit caching. • Token limits & unsupported parameters: • Input token limit is 1,048,576 tokens, and single-segment output token limit is 65,536 tokens for both streaming and non-streaming requests. Firebase AI Logic does not yet support the "long decode" capability. • Sampling parameters (temperature, topP, topK) are ignored, and frequencyPenalty, presencePenalty, and candidateCount fail with a 400 error. • Additional document input types: • In addition to PDFs and plain text, gemini-4-neon (along with other Gemini 4.x models) supports Microsoft Office and OpenDocument files (.doc, .docx, .dotx, .ppt, .pptx, .xls, .xlsx, .odt, .odp, .ods).
Original source - Nov 17, 2026
- Date parsed from source:Nov 17, 2026
- First seen by Releasebot:Nov 18, 2026
- Modified by Releasebot:Nov 18, 2026
November 17, 2026
Firebase releases C++ SDK 14.2.0 with new Authentication features, updates to App Check and Realtime Database, and bug fixes across several services.
SDK Releases
The Firebase SDK for C++ (v14.2.0) is now available. This release includes new features in Authentication; changes to App Check and Realtime Database; bug fixes in Realtime Database, Cloud Functions, and Cloud Storage; and other changes and fixes.
Original source All of your release notes in one feed
Join Releasebot and get updates from Google and hundreds of other software products.
- Nov 16, 2026
- Date parsed from source:Nov 16, 2026
- First seen by Releasebot:Oct 14, 2026
- Modified by Releasebot:Nov 19, 2026
November 16, 2026
Firebase releases updated Flutter, Apple, and JavaScript SDKs, with the Flutter update bringing product changes and bug fixes.
SDK Releases The Firebase SDK for Flutter (v5.2.0) is now available. This release includes changes in several products, and bug fixes in several products. The Firebase Apple SDK (v13.2.0) is now available. For more details, see the Firebase Apple SDK release notes. To install this SDK, see Add Firebase to your app. The Firebase JavaScript SDK (v13.4.0) is now available. For more details, see the Firebase JavaScript SDK release notes. To install the SDK, see Add Firebase to your JavaScript Project.
Original source - Nov 12, 2026
- Date parsed from source:Nov 12, 2026
- First seen by Releasebot:Nov 12, 2026
- Modified by Releasebot:Nov 19, 2026
November 12, 2026
Firebase adds Crashlytics support for web apps in preview and brings the gemini-4-hydrogen model to Firebase AI Logic, expanding app error monitoring and access to newer AI capabilities.
Crashlytics Firebase Crashlytics now supports web apps (in Preview). You can automatically capture browser exceptions and framework errors for Angular, Next.js, and React, upload source maps for de-obfuscated stack traces, customize crash reports with handled exceptions and custom keys, configure alerts, and debug issues using Crashlytics MCP tools. To get started, see Get started with Crashlytics for web apps. Firebase AI Logic Firebase AI Logic supports the latest Gemini stable model gemini-4-hydrogen. You can directly access this model in your mobile or web app via either the Gemini Developer API (free tier) or the Agent Platform Gemini API (formerly Vertex AI). • Short-term availability: For the Agent Platform Gemini API (formerly Vertex AI), gemini-4-hydrogen is a short-term availability model. To avoid service disruptions when models are updated, we strongly recommend using Firebase Remote Config or server prompt templates to dynamically manage the model name in your app. • If you're upgrading from gemini-3.8-flash (or other Gemini 3.x models), note the following differences for gemini-4-hydrogen: • Thinking levels (HIGH default, no XHIGH or thinkingBudget): • Default thinking level is HIGH (changed from MEDIUM on gemini-3.8-flash), which may increase latency and token usage unless you explicitly set thinkingLevel to MEDIUM or LOW. • Supported thinking levels in Firebase AI Logic are LOW, MEDIUM, and HIGH (default). Note that Firebase AI Logic does not currently support XHIGH. • Like gemini-3.8-flash, MINIMAL is not supported (the request will fail with a 400 error). • Unlike gemini-3.8-flash, thinkingBudget is not supported (there's no backwards-compatibility; the request will fail with a 400 error). • Context caching: • Supports implicit caching by default (with a minimum of 4,096 input tokens on the Gemini Developer API and 6,144 input tokens on the Agent Platform Gemini API (formerly Vertex AI)). • Unlike gemini-3.8-flash, gemini-4-hydrogen does not support explicit caching. • Token limits & unsupported parameters: • Input token limit (1,048,576) and single-segment output token limit (65,536 for both streaming and non-streaming requests) match gemini-3.8-flash. Firebase AI Logic does not yet support the "long decode" capability. • Like gemini-3.8-flash, sampling parameters (temperature, topP, topK) are ignored, and frequencyPenalty, presencePenalty, and candidateCount fail with a 400 error. • Additional document input types: • In addition to PDFs and plain text, gemini-4-hydrogen (along with other Gemini 4.x models) supports Microsoft Office and OpenDocument files (.doc, .docx, .dotx, .ppt, .pptx, .xls, .xlsx, .odt, .odp, .ods).
Original source - Nov 11, 2026
- Date parsed from source:Nov 11, 2026
- First seen by Releasebot:Nov 13, 2026
November 11, 2026
Firebase CLI improves Hosting setup and Authentication emulator startup, adds Crashlytics filtering and App Check token persistence, and updates SQL Connect tooling.
Firebase CLI (v15.36.0)
The latest Firebase CLI (v15.36.0) is now available. This version ensures the Firebase Hosting API is enabled during setup, adds key filtering to the Crashlytics MCP list_events tool, improves Authentication emulator startup time, updates the Firebase SQL Connect local toolkit to v3.4.26, and adds local config persistence for App Check debug tokens.
To use the Firebase CLI in your development environment, set up or update the CLI.
Original source Similar to Firebase with recent updates:
- Gemini updates476 release notes · Latest Nov 18, 2026
- Antigravity updates92 release notes · Latest Nov 18, 2026
- Gemini API updates159 release notes · Latest Nov 18, 2026
- Gemini CLI updates174 release notes · Latest Nov 19, 2026
- ChatGPT updates254 release notes · Latest Nov 18, 2026
- Claude updates194 release notes · Latest Nov 17, 2026
- Nov 11, 2026
- Date parsed from source:Nov 11, 2026
- First seen by Releasebot:Nov 12, 2026
- Modified by Releasebot:Nov 13, 2026
November 11, 2026
Firebase adds stable Gemini text-to-speech models with rich vocal direction, broad voice and language options, and streaming audio support. Updated SDKs across Android, JavaScript, and Admin platforms also bring App Check replay protection, token verification, and fixes.
Firebase AI Logic
Firebase AI Logic supports the latest stable Gemini TTS models (gemini-3.8-flash-tts and gemini-3.8-flash-lite-tts) for text-to-speech (TTS) generation via either the Gemini Developer API or the Agent Platform Gemini API (formerly Vertex AI).
- Vocal direction and multi-speaker dialogue: Guide vocal delivery using turn-level SpeechMetadata (style and speaker on TextPart), inline angle-bracket vocal tags (<...>), and pipe backchannels (|...|) for two-speaker conversations.
- Voices and languages: Choose from 30 core prebuilt voices, over 2,000 preset voices from the Extended Voice Library, or custom voices created with Voice design and Voice replication. The models automatically detect the language of the input text across over 100 languages.
- Audio output formats: Unary requests (generateContent) return a complete WAV file (audio/wav), while streaming requests (generateContentStream) return headerless raw 16-bit PCM chunks (audio/l16).
- Migration from gemini-3.1-flash-tts-preview: The preview model gemini-3.1-flash-tts-preview will shut down on November 17, 2026. If your app uses this legacy model, see the migration guide to update your model name, prompt format, and audio handling before that date.
SDK Releases
The Firebase Android SDKs for Firebase AI Logic, App Check, Cloud Firestore, Cloud Messaging, Crashlytics, and Crashlytics NDK 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 v35.2.0. Learn how to add SDKs to your app using the Firebase BoM.
An updated version of the google-services Gradle plugin for Android (v4.6.0) is now available. For more information, see the Firebase Android SDK Release Notes.
The Firebase JavaScript SDK (v13.3.0) 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 (v14.5.1) is now available. This release includes fixes in the core SDK and Authentication. To install this SDK, see Add the Firebase Admin SDK to your server.
The Firebase Admin Java SDK (v9.12.0) is now available. This release adds token verification and replay protection support in App Check, and includes fixes in the core SDK and Authentication. To install this SDK, see Add the Firebase Admin SDK to your server.
The Firebase Admin Python SDK (v7.8.0) is now available. The Admin SDK for Python now requires Python 3.10 or higher and deprecates support for Python 3.10. This release also includes a fix in Authentication. To install this SDK, see Add the Firebase Admin SDK to your server.
The Firebase Admin Go SDK (v4.23.0) is now available. The Admin SDK for Go now requires Go 1.26 or higher. This release also adds replay protection support in App Check and includes a fix in Authentication. To install this SDK, see Add the Firebase Admin SDK to your server.
The Firebase Admin .NET SDK (v3.7.1) is now available. This release includes fixes in Authentication and Cloud Messaging. To install this SDK, see Add the Firebase Admin SDK to your server.
Original source - Nov 10, 2026
- Date parsed from source:Nov 10, 2026
- First seen by Releasebot:Nov 11, 2026
November 10, 2026
Firebase AI Logic now supports the stable Gemini 4 Carbon model in mobile and web apps through either Gemini API provider, with Blaze billing required. The release also outlines model limits and unsupported settings to help developers configure requests and manage updates.
Firebase AI Logic
Firebase AI Logic supports the latest Gemini stable model gemini-4-carbon. You can directly access this model in your mobile or web app via either the Gemini Developer API or the Agent Platform Gemini API (formerly Vertex AI).
- Billing: Accessing gemini-4-carbon requires the pay-as-you-go Blaze pricing plan regardless of your Gemini API provider.
- Short-term availability: For the Agent Platform Gemini API (formerly Vertex AI), gemini-4-carbon is a short-term availability model. To avoid service disruptions when models are updated, we strongly recommend using Firebase Remote Config or server prompt templates to dynamically manage the model name in your app.
- If you're migrating from previous models, note the following for this new model:
- Token limits:
- Default and maximum output token limits: 32,768 tokens for non-streaming requests and 65,536 tokens for streaming requests. The input token limit is the same as the older Gemini 3.x models.
- Firebase AI Logic does not yet support the "long decode" capability.
- No support for MINIMAL thinking level or thinkingBudget:
- Supports the following thinking levels: LOW, MEDIUM (default), and HIGH.
- Does not support setting thinking level of MINIMAL. The request will fail with a 400 error.
- Does not support using thinkingBudget (there's no backwards-compatibility). The request will fail with a 400 error.
- Context caching:
- Supports implicit caching by default (with a minimum of 4,096 input tokens on the Gemini Developer API and 6,144 input tokens on the Agent Platform Gemini API (formerly Vertex AI)).
- Does not support explicit caching.
- Unsupported parameters: Does not support temperature, top-K, top-P, frequency penalty, presence penalty, or candidate count.
- Temperature, top-K, or top-P: Ignored by the model.
- Frequency penalty, presence penalty, and candidate count: The request will fail with a 400 error.
For more deterministic responses, define system instructions with explicit rules for your specific use case.
- Prefilled model turn validation: Does not support prefilling model turns. If the last non-empty turn in the request is a model turn, the request will fail with a 400 error.
- Token limits:
- Nov 9, 2026
- Date parsed from source:Nov 9, 2026
- First seen by Releasebot:Nov 14, 2026
- Modified by Releasebot:Nov 14, 2026
November 09, 2026
Firebase releases GA support for the drop() command in Cloud Firestore with MongoDB compatibility.
Cloud Firestore
Support for the
Original sourcedrop()command for Firestore with MongoDB compatibility is generally available (GA). - Nov 4, 2026
- Date parsed from source:Nov 4, 2026
- First seen by Releasebot:Nov 5, 2026
November 04, 2026
Firebase CLI improves SQL Connect and Emulator Suite behavior on Linux and Windows, and updates the Crashlytics MCP get_report tool.
The latest Firebase CLI (v15.35.0) is now available. This version fixes query latency on Linux in the SQL Connect PGlite emulator, prevents extra console windows on Windows when starting Java-based emulators, fixes listen address display for the Emulator Suite UI, and updates the Crashlytics MCP get_report tool.
To use the Firebase CLI in your development environment, set up or update the CLI.
Original source - Nov 3, 2026
- Date parsed from source:Nov 3, 2026
- First seen by Releasebot:Nov 4, 2026
November 03, 2026
Firebase adds an analytics skill to help AI assistants set up Firebase and measure screenviews on iOS and Android.
Analytics
The new firebase-analytics skill is now available in Firebase agent skills to help AI assistants with getting started (base setup and SDK initialization) and measuring screenviews for Google Analytics on iOS and Android.
Original source - Nov 2, 2026
- Date parsed from source:Nov 2, 2026
- First seen by Releasebot:Nov 7, 2026
November 02, 2026
Firebase will automatically route Gemini 2.5 Pro and Flash API requests to newer Gemini 3.x models starting December 1, 2026. The change brings updated model behavior and billing, with unsupported parameters and request formats requiring developers to review and test configurations.
Gemini API
For the Gemini Developer API (Google AI), starting December 1, 2026, requests to the Gemini 2.5 Pro and Flash models will be automatically routed to newer Gemini 3.x models:
- Gemini 2.5 Pro (gemini-2.5-pro): Requests will be automatically routed to gemini-3.8-flash.
- Gemini 2.5 Flash (gemini-2.5-flash): Requests will be automatically routed to gemini-3.5-flash-lite.
- Gemini 2.5 Flash‑Lite (gemini-2.5-flash-lite): No changes at this time.
Note that for the Agent Platform Gemini API (formerly Vertex AI), gemini-2.5-pro, gemini-2.5-flash, and gemini-2.5-flash-lite were already shut down (retired) on October 20, 2026.
Breaking changes and behavioral differences when requests are routed to Gemini 3.x models
- Billing: Inference will be billed at the rates of the newer models.
- Unsupported parameters: Both gemini-3.8-flash and gemini-3.5-flash-lite do not support several sampling parameters:
- Frequency penalty, presence penalty, and candidate count (for gemini-2.5-pro routed to gemini-3.8-flash): gemini-3.8-flash does not support frequencyPenalty, presencePenalty, or candidateCount. Requests that include these parameters will fail with a 400 error.
- Temperature, top-K, and top-P: All general-use Gemini 3.x models ignore temperature, topK, and topP. For more deterministic responses, define system instructions with explicit rules for your specific use case.
- Prefilled model turn validation: Prefilling model turns is no longer supported in Gemini 3.x models. If the last non-empty turn in the request is a model turn, the request will fail with a 400 error.
- Thinking configuration:
- No support for MINIMAL thinking level on gemini-3.8-flash: If you set the thinking level to MINIMAL for a request routed to gemini-3.8-flash, the request will fail with a 400 error.
- Use thinking levels instead of thinking budgets: Thinking budgets are supported by Gemini 3.x models for backwards-compatibility, but they are strongly discouraged (especially starting with gemini-3.8-flash). We recommend migrating from thinkingBudget to thinkingLevel.
If you still use Gemini 2.5 models with Firebase AI Logic, we strongly recommend explicitly updating your app to use a newer model (like gemini-4-argon, gemini-3.8-flash, or gemini-3.5-flash-lite) and testing your model configuration before December 1, 2026. For more details, see the deprecated and shutdown models FAQ.
Original source - Oct 30, 2026
- Date parsed from source:Oct 30, 2026
- First seen by Releasebot:Nov 5, 2026
- Modified by Releasebot:Nov 19, 2026
October 30, 2026
Firebase adds custom constraints for Cloud Storage policies and releases JavaScript SDK v13.2.0.
Cloud Storage for Firebase Custom constraints for Cloud Storage for Firebase are now available. You can use custom constraints to enforce policies on Cloud Storage resources, such as a policy that enforces all buckets to have Object Versioning enabled. SDK Releases The Firebase JavaScript SDK (v13.2.0) is now available. For more details, see the Firebase JavaScript SDK release notes. To install the SDK, see Add Firebase to your JavaScript Project.
Original source - Oct 30, 2026
- Date parsed from source:Oct 30, 2026
- First seen by Releasebot:Jul 31, 2026
- Modified by Releasebot:Nov 4, 2026
October 30, 2026
Firebase releases JavaScript SDK v13.2.0, with details and installation guidance available in its release notes.
SDK Releases
The Firebase JavaScript SDK (v13.2.0) is now available. For more details, see the Firebase JavaScript SDK release notes. To install the SDK, see Add Firebase to your JavaScript Project.
Original source - Oct 28, 2026
- Date parsed from source:Oct 28, 2026
- First seen by Releasebot:Nov 2, 2026
October 28, 2026
Firebase CLI fixes Firestore emulator rule watching and the MCP server environment command, updates SQL Connect tooling, and includes dependency and security updates.
Firebase CLI (v15.34.2)
The latest Firebase CLI (v15.34.2) is now available. This version fixes an issue with the Cloud Firestore emulator watching rules files, fixes the Firebase MCP server environment command, updates the Firebase SQL Connect local toolkit to v3.4.25, and includes dependency and security updates.
To use the Firebase CLI in your development environment, set up or update the CLI.
Original source - Oct 27, 2026
- Date parsed from source:Oct 27, 2026
- First seen by Releasebot:Nov 2, 2026
October 27, 2026
Firebase AI Logic now lets teams temporarily delay App Check enforcement ahead of automatic protection planned for November 2026.
Firebase AI Logic
You can now temporarily delay the automatic enforcement of Firebase App Check for Firebase AI Logic.
Starting on November 2, 2026, Firebase will automatically enforce App Check for Firebase AI Logic in your project (if it isn't already) to help protect the Gemini API. If you have released app versions that don't yet have App Check implemented, you can temporarily delay auto-enforcement.
Learn more about this upcoming auto-enforcement of App Check for Firebase AI Logic.
Original source
Curated by the Releasebot team
Releasebot is an aggregator of official product update announcements from hundreds of software vendors and thousands of sources.
Our editorial process involves the manual review and audit of release notes procured with the help of automated systems.