Gemini Releases
Google Gemini product news and SDK releases. New features, breaking changes, security advisories and deprecations - each summarised in plain English and updated continuously.
Tracking 148 Gemini releases · Updated
- endoflife.date eolgemini-models ·
Gemini model gemini-2.5-flash-image reaches end of life in 7 days
Gemini model gemini-2.5-flash-image reaches end of life on 2026-10-02 (7 days from now).
deprecation - Genkit Releases sdkaipreviewengineermedia ·
Genkit Python SDK v0.12.0 Enhances AI Model Integration and Response Handling
Genkit Python SDK v0.12.0 introduces A2UI surfaces, integrates new Google AI models like Deep Research and Lyria, and strengthens the OpenAI plugin with support for `gpt-6-astra`. This release refines `generate()` error handling to return `ModelResponse` even when a turn fails, enabling better retry mechanisms for multi-turn interactions. Additionally, it streamlines the `Part` API for message content and moves agents to an experimental namespace. These updates are designed to help developers build more interactive and robust generative AI applications.
patch - Gemini Blog aigaarchitect ·
Gemini 3.8 Live Introduces Real-Time Live Avatar for Enterprise
Google has launched Live Avatar for Gemini 3.8 Live, bringing near real-time visual presence to its native live dialogue models. This feature pairs real-time video generation with speech, enabling dynamic visual personas for interactive virtual offerings. It enhances customer service and walkthroughs by supporting precise lip-syncing, natural expressions, and fluid turn-taking. Live Avatar is now available in Gemini Enterprise, offering multimodal conversations, asynchronous tool execution, multilingual support across 97 languages, and custom avatar creation.
feature - Gemini Blog aigaarchitect ·
Gemini expands Connected Apps integrations with new wave of services
Google's Gemini is rolling out a new wave of Connected Apps, allowing users to integrate popular productivity, creativity, and lifestyle tools directly into the AI assistant. This update streamlines workflows by reducing the need to switch between applications, enabling users to manage projects, design assets, and plan activities within Gemini. The rollout, beginning today, includes services like Airtable, Adobe, monday.com, Peloton, and apartments.com. Users can connect these apps through Gemini settings or by using @mentions in chat.
feature - Gemini Blog aigaengineer ·
Gemini 3.8 Flash and Flash-Lite TTS Models Offer Advanced Voice Generation
Google has introduced two new text-to-speech models, Gemini 3.8 Flash TTS and 3.8 Flash-Lite TTS, designed to transform voice generation into a dynamic creative studio. These models allow creators, developers, and enterprises to create bespoke voices, customize roles, accents, and characteristics, and direct performance line by line with granular control. Key features include generative voice design, an expansive voice library, and voice replication with built-in consent verification and SynthID watermarking. The models are rolling out starting today for developers via the Gemini API and Google AI Studio, with enterprise availability coming soon.
feature - endoflife.date eolgemini-models ·
Gemini model gemini-omni-flash-preview reaches end of life in 7 days
Gemini model gemini-omni-flash-preview reaches end of life on 2026-09-30 (7 days from now).
deprecation - Gemini Python SDK Releases sdkaigaengineer ·
Python GenAI SDK v2.25.0 enhances Voice and Speech APIs
Version 2.25.0 of the `googleapis/python-genai` SDK introduces new Voice and Speech APIs, including a dedicated `Voices` API resource and `sample_audio` functionality within the Voice feature. These updates provide developers with enhanced capabilities for integrating advanced speech generation and configuration into their AI applications. The release also exposes key metadata and configuration options such as `SpeechMetadata`, `VoiceConfig.voice`, and `SpeechAnnotation` in the public SDKs. Additionally, the update includes bug fixes, such as sanitizing path parameter descriptions in GAOS OpenAPI generation.
feature patch - Genkit Releases sdkaigaengineer ·
Genkit CLI 1.43.0 Enhances Logging and Developer UI Capabilities
Genkit CLI version 1.43.0 introduces new features and fixes, focusing on improved debugging and developer experience for AI agents. It adds a `genkit log:list` command for querying structured runtime logs and OTLP log ingestion to stream Python debug logs to the Developer UI. The Developer UI now features a top-level trace logs panel, supports zip archive downloads for multi-part agent artifacts, and offers improved resume behavior for human-in-the-loop agent interactions. These updates streamline the process of monitoring and interacting with AI applications built using Genkit.
feature patch - Gemini Python SDK Releases sdkaigaengineer ·
Google's Python GenAI SDK v2.24.0 Adds Credential APIs and Retrieval Steps
Version 2.24.0 of the Google Python GenAI SDK introduces new credential APIs and extends the interactions schema with RetrievalCallStep and RetrievalResultStep. This update enhances capabilities for managing authentication and integrating retrieval-augmented generation features within AI applications. Python developers leveraging the GenAI SDK for Gemini models will benefit from these additions, alongside improvements for environment file handling. The release also includes a bug fix for automatic function calling budget management and documentation updates.
feature patch - Gemini Blog aigaengineer ·
Google introduces Gemini 3.8 Live models for advanced real-time voice AI
Google has introduced Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, two new models designed to enhance near real-time reasoning for voice agents. These models aim to make AI conversations more intuitive, intelligent, and fluid by combining conversational intelligence with visual grounding and multi-step reasoning. They are available today for developers via the Gemini API and Google AI Studio, with private previews for enterprises and broader rollout for Google Workspace and Search users. The release empowers developers and enterprises to build production-ready voice-driven interfaces.
feature - Gemini Blog ai ·
Gemini AI helps tackle household chores with new productivity applications
Google highlights four practical ways Gemini can assist with managing household tasks, extending its utility beyond professional settings. The AI can troubleshoot DIY repairs using camera input, streamline meal planning with personalized recipes and Instacart integration, and track household spending by extracting information from receipt photos. Additionally, Gemini can help compare complex purchases like cars, providing side-by-side comparisons and negotiation tips.
announcement - Gemini Blog aiengineer ·
Google Developer Groups Announce DevFest 2026 for Hands-On AI Exploration
Google Developer Groups (GDGs) have announced the return of DevFest for its 2026 season, running from October 1 to December 31, 2026. The global, community-led conferences will gather nearly a million developers to focus on hands-on experimentation with Google's AI technologies, including Gemini and Google AI Studio. The theme, "Build, Secure, Scale: Developers and Builders in the Agentic Era," emphasizes secure deployment, data privacy, and responsible AI. DevFest events are tailored by local GDGs to address specific community needs and foster collaborative skill development.
announcement - Gemini Python SDK Releases sdkaiengineer ·
Google Python GenAI Library v2.23.0 Adds Environment Copying and Dynamic Turn Completion
Version 2.23.0 of the Google Python GenAI library introduces new capabilities for developers using generative AI models. It adds `from_environment` support for environment copying, which streamlines configuration management. The update also facilitates dynamic turn completion via `interaction_status` in `live.py receive()`, enhancing real-time interaction control within applications. Additionally, a minor bug fix addresses an unused field in `FunctionResultDelta`, improving code stability.
feature patch - Gemini Blog aigaengineer ·
Gemini Desktop App Now Available for Windows Users
Google has launched a dedicated Gemini app for Windows 10 and 11, enabling instant AI assistance directly from the desktop. The app allows users to access Gemini with a keyboard shortcut, manage multi-step tasks in a dedicated workspace, and generate images and videos. Designed for seamless integration into daily workflows, it runs quietly without significant performance impact. The app is globally available starting today, with plans for more native desktop capabilities to be rolled out over time.
feature - Gemini Blog aiengineer ·
4 ways Gemini makes administrative chores quick and easy
The provided article focuses on how Gemini can streamline administrative tasks, suggesting four distinct approaches for making these chores quicker and easier. The accompanying visual depicts an "Ask Gemini" prompt box alongside a person reviewing documents, illustrating a potential user interaction with Gemini for administrative support.
announcement - Gemini Blog aigaengineer ·
Lyria 3.5 music generation model generally available in Gemini
Google has made its Lyria 3.5 music generation model broadly available within the Gemini app and via the Gemini API. This update enhances music creation with more expressive vocals and richer arrangements, enabling users to produce higher fidelity tracks. Within the Gemini app, users gain new capabilities like genre selection, vocal/instrumental style choice, creative templates, and track length specification. Lyria 3.5 is accessible globally to all users, artists on Google Flow Music, and developers through Google AI Studio and Google Vids.
feature announcement - Genkit Releases sdkaipreviewengineermediaeducation ·
Genkit Python SDK v0.11.0 Improves Error Handling, Video Integration, and Model Configuration
The Genkit Python SDK v0.11.0 enhances developer experience by preventing loss of model replies during certain `generate()` failures, allowing retries from successful tool rounds. It introduces a new asynchronous `Veo` model for video generation via polling operations, treating video as a background job. Additionally, the update unifies provider HTTP errors into a single `GenkitError` type, simplifying error handling and retries across different AI models. Developers can now also use typed model references, ensuring configuration compatibility and better auto-completion for `generate()` calls.
patch - Genkit Releases sdkaipreviewengineer ·
Genkit Go v1.13.1 Retracts v1.13.0 and Corrects A2UI Plugin Path
Genkit Go v1.13.1 has been released, retracting the prior v1.13.0 due to a package path discrepancy. This version corrects the A2UI preview plugin's path to `a2ui/exp`, aligning it with documentation. The retraction ensures `go get` commands resolve correctly and hide the problematic v1.13.0. Developers using or planning to use the A2UI plugin must update their import paths for this change.
breaking patch - Genkit Releases sdkaipreviewengineer ·
Genkit Go v1.13.0 Adds Resilient AI Agents & A2UI Interactive UI Plugin Preview
Genkit Go v1.13.0 introduces significant improvements to AI agent resilience and error handling. The `Generate` function now returns partial responses and classified errors on failure or abort, enabling developers to retry failed steps without repeating successful tool calls. Agents can commit failed or aborted conversations as resumable snapshots, enhancing recovery and user experience, while sub-agents can run in the background. Additionally, an experimental A2UI plugin is available in preview, allowing agents to stream interactive UI to a browser.
feature - Gemini Python SDK Releases sdkaigaengineer ·
Google's Python GenAI SDK adds support for Gemini 3.8 Flash model
The Google Python GenAI SDK, updated to version 2.22.0, now includes direct support for the Gemini 3.8 Flash model. This enhancement allows developers to easily integrate the new, efficient large language model into their applications using the SDK. The release also incorporates updated descriptions for existing Flash models. This provides engineers with tools to build more performant generative AI solutions.
feature
About Gemini release tracking on ReleaseBytes
Google's Gemini models and APIs move fast: new model versions, pricing changes, SDK releases and capability launches arrive weekly. ReleaseBytes tracks the official Gemini and Google DeepMind announcement channels and SDK changelogs, so you can see every model update, deprecation notice and API change in one feed.
Frequently asked questions
How often are Gemini release notes updated on ReleaseBytes? ›
Continuously. ReleaseBytes monitors the official Gemini release channels around the clock and publishes a plain-English summary of each announcement shortly after it lands.
What kinds of Gemini changes does ReleaseBytes track? ›
New features, enhancements, bug fixes, security advisories, breaking changes, deprecations and end-of-life announcements. Every item is tagged by type so you can filter to just the changes that need action.
How can I get alerts for new Gemini releases? ›
Set up a free email or Slack alert filtered to Gemini, subscribe to the weekly digest, or follow the RSS feed. Teams can also install the ReleaseBytes GitHub App or connect via MCP.
Where does the Gemini release data come from? ›
From the official sources: Google Gemini product news and SDK releases. Every item links back to the original vendor announcement.