Supervised fine-tuning for Gemini 3.8 Flash is available in General Availability
Supervised fine-tuning, continuous tuning, and tuning checkpoints for Gemini 3.8 Flash are now generally available on us and eu multi-region endpoints.
Also shipped Nov 20, 2026Google Cloud in November 2026 →
Improvements on the pull request overview in public preview
GitHub updated the pull request overview page in public preview, introducing clearer merge status indicators, collapsible reviews and comments, and persistent branch visibility on smaller screens.
Cloudflare Images - View delivery analytics for hosted images
Cloudflare introduced an Images Analytics dashboard for Cloudflare Images, offering up to 30 days of delivery metric trends and SQL access via the cf sql query CLI tool.
Security advisory permissions are now generally available
GitHub has made security advisory permissions generally available, allowing organization owners to assign specific security-related roles to users without granting full administrative access.
Map sensitive data with Datadog Data Security Posture Management
Datadog Data Security Posture Management (DSPM) is now in preview, enabling users to identify PII, financial data, and secrets at rest in Snowflake, PostgreSQL, and Amazon S3.
Retirement: Support for Azure Functions OpenAI Extension ends on April 20, 2027
Support for the Azure Functions OpenAI extension will end on April 20, 2027. Users are advised to migrate to the OpenAI, Microsoft Foundry, or Azure AI Search SDKs.
[In preview] Public Preview: Programmed Training in Microsoft Foundry
Programmed Training, an API for building custom post-training loops over open-weight models, is now available in public preview in Microsoft Foundry.
More from Gemini Enterprise Agent PlatformEvery Gemini Enterprise Agent Platform release →
Gemini 4 Hydrogen in CodeMender (Preview)
Google Cloud added preview support for Gemini 4 Hydrogen in CodeMender, accessible by supplying the --model gemini-4-hydrogen flag to cm subcommands.
Gemini 4 Neon (GA)
Google Cloud announced the general availability of Gemini 4 Neon in Gemini Enterprise Agent Platform, supporting a 1M-token context window and four thinking levels.
AlphaGenome available without an allowlist
Google Cloud removed project allowlisting requirements for AlphaGenome models alphagenome-003 and alphagenome-004 in Model Garden.
Deferred tier for asynchronous model calls on the Interactions API
Google Cloud announced general availability for the Deferred tier on the Gemini Enterprise Agent Platform. It offers a 50% discount compared to Standard PayGo rates for asynchronous Gemini model calls, providing a 24-hour completion window and 7-day state retention.
DeepSeek-V4.1-Flash-1017 is available in Public Preview
DeepSeek-V4.1-Flash-1017 is available in public preview as a fully managed model in Vertex AI Model Garden. The updated model features improved reasoning, coding, agentic, and visual capabilities, supporting multimodal inputs with up to 1M context tokens.
Deferred tier for autonomous agent scheduling is generally available
Autonomous agent scheduling (the deferred tier ) is available in General Availability (GA) and now supports Agent Development Kit (ADK) agents ( google-adk v2.10.0 and later) in addition to the Gemini Deep Research Agent .
Voice remixing for Gemini 3.8 TTS models in Preview
Voice remixing is available in Preview for Gemini 3.8 Flash TTS and Gemini 3.8 Flash-Lite TTS . Voice remixing creates a new custom voice from an existing prebuilt, Extended Voice Library, designed, or remixed voice. You provide a natural-language description of the changes to make, such as a different age, pitch, energy, or accent.
Agent Gateway for GKE in Preview
Google Cloud released Agent Gateway for GKE in Preview for clusters on GKE version 1.37.2-gke.1042000 or higher in europe-west1 and us-central1.