This page documents production updates to Vertex AI. Check this page for announcements about new or updated features, bug fixes, known issues, and deprecated functionality.
You can see the latest product updates for all of Google Cloud on the
[Google Cloud](https://docs.cloud.google.com/release-notes.md.txt) page, browse and filter all release notes in the
[Google Cloud console](https://console.cloud.google.com/release-notes),
or programmatically access release notes in
[BigQuery](https://console.cloud.google.com/bigquery?p=bigquery-public-data&d=google_cloud_release_notes&t=release_notes&page=table).

To get the latest product updates delivered to you, add the URL of this page to your
[feed
reader](https://wikipedia.org/wiki/Comparison_of_feed_aggregators), or add the
[feed URL](https://docs.cloud.google.com/feeds/generative-ai-on-vertex-ai-release-notes.xml) directly.

## May 26, 2026

Deprecated **Vertex AI Extensions deprecation**

Vertex AI Extensions is deprecated and will be shut down after November 26, 2026.
We recommend
[migrating to Agent Platform](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/extensions/migrate.md.txt)
to avoid service disruption.

## April 17, 2026

Feature **RAG Cross Corpus Retrieval**

RAG Cross Corpus Retrieval is available in [public preview](https://cloud.google.com/products#product-launch-stages). This feature allows you to retrieve relevant contexts or generate answers from multiple RAG corpora simultaneously using the `AsyncRetrieveContexts` and `AskContexts` APIs.

For more information, see [RAG Cross Corpus Retrieval](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/rag-engine/cross-corpus-retrieval).

## April 14, 2026

Feature **Anthropic's Claude Opus 4.7**

[Claude Opus 4.7](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/partner-models/claude/opus-4-7)
is available in Model Garden.

## April 06, 2026

Feature **Metadata search for RAG Engine**

Use schema-based metadata search in Vertex AI RAG Engine.
You can define a metadata schema for a corpus, attach metadata to files within
that corpus, and use this metadata to filter contexts during retrieval.
For more information, see
[Filter with metadata search](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/rag-engine/use-metadata-search).

## April 03, 2026

Feature **Gemma 4 26B A4B IT** is available as an experimental launch in Model Garden. This is an open model built by Google DeepMind. Gemma 4 models are multimodal, handling text and image input and generating text output.
Gemma 4 26B A4B IT is available as a managed API in Model Garden. To learn more, see
[Gemma 4 26B A4B IT](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/maas/google/gemma-4-26b-a4b-it).
Feature **Vertex AI RAG Engine Serverless mode**

Vertex AI RAG Engine Serverless mode is now available in [public
preview](https://cloud.google.com/products#product-launch-stages). Serverless
mode provides a fully managed database for storing RAG resources that abstracts
away database provisioning and scaling. You can seamlessly switch between
Serverless mode and Spanner mode, which provides dedicated, isolated database
instances.

For more information, see the following:

- [Deployment modes in Vertex AI RAG Engine](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/rag-engine/deployment-modes)
- [Serverless mode](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/rag-engine/serverless-mode)
- [Managing Spanner mode](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/rag-engine/spanner-mode)
- [Switching between modes](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/rag-engine/switching-modes)

## April 02, 2026

Feature **Veo 3.1 Lite**

Veo 3.1 Lite is available in [public
preview](https://cloud.google.com/products#product-launch-stages). This release
is our most cost-efficient Veo on Vertex AI model.

For more information, see [3.1 Lite
Generate](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/veo/3-1-generate#3.1-lite-generate-001)
Announcement **Gemini 2.5 model retirement dates updated**

The retirement dates for Gemini 2.5 Pro, Gemini 2.5 Flash-Lite,
and Gemini 2.5 Flash have been updated to October 16, 2026. For more information,
see [Model versions and lifecycle](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions).

## March 25, 2026

Feature **Lyria 3**

Lyria is available in [public
preview](https://cloud.google.com/products#product-launch-stages). You can use
`lyria-3-pro-preview` to generate 184 seconds of audio, or
`lyria-3-clip-preview` to generate 30 seconds of audio.

For more information, see the following:

- [Lyria 3 Pro
  Preview](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/lyria/lyria-3#lyria-3-pro-preview)

- [Lyria 3 Clip Preview](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/lyria/lyria-3#lyria-3-clip-preview)

## March 24, 2026

Deprecated **Imagen generation GA endpoints deprecation**

The following table describes image generation endpoints that are deprecated and
their replacements. We recommend updating your model endpoints before June 30,
2026, to avoid service disruption.

| Discontinued endpoints | Recommended endpoint migration |
|---|---|
| `imagegeneration@002` | `gemini-2.5-flash-image` |
| `imagegeneration@003` | `gemini-2.5-flash-image` |
| `imagegeneration@004` | `gemini-2.5-flash-image` |
| `imagegeneration@005` | `gemini-2.5-flash-image` |
| `imagegeneration@006` | `gemini-2.5-flash-image` |
| `imagetext@001` | `gemini-2.5-flash-image` |
| `imagen-3.0-capability-001` | `gemini-2.5-flash-image` |
| `imagen-3.0-capability-002` | `gemini-2.5-flash-image` |
| `imagen-3.0-fast-generate-001` | `gemini-2.5-flash-image` |
| `imagen-3.0-generate-001` | `gemini-2.5-flash-image` |
| `imagen-3.0-generate-002` | `gemini-2.5-flash-image` |
| `imagen-4.0-fast-generate-001` | `gemini-2.5-flash-image` |
| `imagen-4.0-generate-001` | `gemini-2.5-flash-image` |
| `imagen-4.0-ultra-generate-001` | `gemini-2.5-flash-image` |

Deprecated **Video generation GA endpoints deprecation**

The following table describes video generation endpoints that are deprecated and
their replacements. We recommend updating your model endpoints before June 30,
2026, to avoid service disruption.

| Discontinued endpoints | Recommended endpoint migration |
|---|---|
| `veo-3.0-generate-001` | `veo-3.1-generate-001` |
| `veo-3.0-fast-generate-001` | `veo-3.1-fast-generate-001` |
| `veo-2.0-generate-001` | `veo-3.1-generate-001` |

## March 12, 2026

Feature **Partner model evaluations**

The Gen AI evaluation service supports evaluating partner models, such as Anthropic
and Llama models. For more information, see
[Perform evaluation using the console](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/evaluation-genai-console#evaluate_partner_models).

## March 03, 2026

Deprecated **Video generation preview endpoints deprecation**

The following table describes video generation endpoints that are deprecated and
their replacements. We recommend updating your model endpoints before April 2,
2026, to avoid service disruption.

| Discontinued endpoints | Recommended endpoint migration |
|---|---|
| `veo-3.0-generate-preview` | `veo-3.0-generate-001` |
| `veo-3.0-fast-generate-preview` | `veo-3.0-fast-generate-preview` |
| `veo-2.0-generate-preview` | `veo-2.0-generate-001` |
| `veo-2.0-generate-exp` | `veo-2.0-generate-001` |
| `veo-001-preview-0815` | `veo-2.0-generate-001` |
| `veo-001-preview` | `veo-2.0-generate-001` |
| `veo-3.1-generate-preview` | `veo-3.1-generate-001` |
| `veo-3.1-fast-generate-preview` | `veo-3.1-fast-generate-001` |

Feature **Gemini 3.1 Flash-Lite**

Gemini 3.1 Flash-Lite (`gemini-3.1-flash-lite`) is
available in [public preview](https://cloud.google.com/products#product-launch-stages).
This release is our most cost-efficient Gemini model and is
optimized for low latency use cases for high-volume, cost-sensitive LLM traffic.

For more information, see
[Gemini 3.1 Flash-Lite](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/3-1-flash-lite).

## February 26, 2026

Feature **Gemini 3.1 Flash Image**

Gemini 3.1 Flash Image (`gemini-3.1-flash-image`) is available
in [public preview](https://cloud.google.com/products#product-launch-stages).
This release enables high-quality image generation with improved pricing and
latency. We recommend using Gemini 3.1 Flash Image when generating
images.

For more information, see
[Gemini 3.1 Flash Image](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/3-1-flash-image).

## February 23, 2026

Deprecated **Anthropic's Claude 3 Haiku**

Anthropic's Claude 3 Haiku is deprecated as of February 23, 2026 and will be
shut down on August 23, 2026. For more information, see
[Partner model deprecations](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/deprecations/partner-models#haiku-3).

## February 19, 2026

Feature **Gemini 3.1 Pro Preview**

[Gemini 3.1 Pro](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/3-1-pro)
is available in preview in Model Garden. Gemini 3.1 Pro is
our most advanced reasoning Gemini model, capable of solving complex
problems from different information sources, including text, audio, images,
video, PDFs, and even entire code repositories with its 1M token context
window.

## February 17, 2026

Feature **Anthropic's Claude Sonnet 4.6**

[Claude Sonnet 4.6](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/partner-models/claude/sonnet-4-6)
is available in Model Garden.
Deprecated **Image generation preview endpoints deprecation**

The following table describes image generation endpoints that are deprecated and
their replacements. We recommend updating your model endpoints before March 19,
2026, to avoid service disruption.

| Discontinued endpoints | Recommended endpoint migration |
|---|---|
| `gemini-2.0-flash-image-generation-preview` | `gemini-2.5-flash-image` |
| `gemini-2.5-flash-image-generation-preview` | `imagen-4.0-generate-001` or `gemini-2.5-flash-image` |
| `imagen-4.0-generate-preview-05-20` | `imagen-4.0-generate-001` or `gemini-2.5-flash-image` |
| `imagen-4.0-generate-preview-06-06` | `imagen-4.0-generate-001` or `gemini-2.5-flash-image` |
| `imagen-4.0-ultra-generate-preview-06-06` | `imagen-4.0-generate-001` or `gemini-2.5-flash-image` |
| `imagen-4.0-fast-generate-preview-05-20` | `imagen-4.0-generate-001` or `gemini-2.5-flash-image` |
| `imagen-product-recontext-preview-06-30` | `gemini-2.5-flash-image` |
| `imagen-2.0-edit-preview-0627` | `gemini-2.5-flash-image` |
| `virtual-try-on-preview-08-04` | `virtual-try-on-001` |
| `imagen-4.0-ingredients-preview` | `gemini-2.5-flash-image` |

## February 10, 2026

Feature **GLM 5** is available as an experimental launch in Model Garden. This model
is targeting complex systems engineering and long-horizon agentic tasks.
GLM 5 is available as a managed API in Model Garden. To learn more, see
[GLM 5](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/maas/zai-org/glm-5).

## February 04, 2026

Feature **Anthropic's Claude Opus 4.6**

[Claude Opus 4.6](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/partner-models/claude/opus-4-6)
is available in Model Garden.

## January 23, 2026

Feature **Virtual Try-On**

Virtual Try-On is now [generally available
(GA)](https://cloud.google.com/products/#product-launch-stages).
The new endpoint, `virtual-try-on-001`,
replaces the previous endpoint, `virtual-try-on-preview-08-04`. We
recommend changing to the new endpoint as soon as possible.

For more information, see
[Generate Virtual Try-On Images](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/generate-virtual-try-on-images.md.txt).

## January 22, 2026

Announcement Codestral (25.01) and Mistral Large (24.11) are retired as of January 23, 2026.

## January 20, 2026

Feature **GLM 4.7** GA is now available in Model Garden. This model
is designed for core or vibe coding, tool use, and complex reasoning.
GLM 4.7 is available as a managed API in Model Garden. To learn more, see
[GLM 4.7](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/maas/zai-org/glm-47).

## January 13, 2026

Feature **Veo 3.1 reference-to-video update**

Veo 3.1 Preview models now support the following features:

- 9:16 aspect ratio for reference-to-video.
- Upsampling for videos generated at 1080p and 4k resolutions.

For more information, see the following:

- [Generate Veo videos from reference images](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/use-reference-images-to-guide-video-generation)
- [Veo on Vertex AI video generation API](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo-video-generation)

## January 06, 2026

Feature **GLM 4.7** is available as an experimental launch in Model Garden. This model
is designed for core or vibe coding, tool use, and complex reasoning.
GLM 4.7 is available as a managed API in Model Garden. To learn more, see
[GLM 4.7](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/maas/zai-org/glm-47).

## January 05, 2026

Deprecated **Anthropic's Claude 3.5 Haiku**

Anthropic's Claude 3.5 Haiku is deprecated as of January 5, 2026 and will be
shut down on July 5, 2026. For more information, see
[Partner model deprecations](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/deprecations/partner-models#haiku-3-5).

## December 18, 2025

Feature The following models are available through Model Garden:

- [FunctionGemma](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/functiongemma)
- [T5Gemma 2](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/t5gemma)
Feature **Save and share prompts in Vertex AI Studio** : The prompt sharing feature no longer needs to be enabled. You can share prompts without asking your administrator to first enable the prompt sharing feature. For more information, see [Save and share prompts](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/prompt-sharing).

## December 17, 2025

Feature [Cloud API Registry](https://docs.cloud.google.com/api-registry/docs/overview.md.txt) is available
in the Google Cloud console in
[Preview](https://cloud.google.com/products#product-launch-stages). Use
Cloud API Registry in the Google Cloud console to view and manage the MCP
servers and tools your agent has access to.
Feature **Gemini 3 Flash**

Gemini 3 Flash is now available in public preview. This model is designed to
tackle the most challenging agentic problems with strong coding and
state-of-the-art reasoning capabilities, and is our best model for complex
multimodal understanding.

For more information, see [Gemini 3
Flash](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/3-flash).

## December 16, 2025

Change Updated pricing for **Vertex AI Agent Engine**:

- Pricing for Vertex AI Agent Engine Runtime was lowered.

- On **January 28, 2026**, Sessions, Memory Bank, and Code Execution will begin
  charging for usage.

For more information, see [Pricing](https://cloud.google.com/vertex-ai/pricing#vertex-ai-agent-engine).
Announcement **Vertex AI Agent Engine**

Vertex AI Agent Engine [Sessions](https://docs.cloud.google.com/agent-builder/agent-engine/sessions/overview)
and [Memory Bank](https://docs.cloud.google.com/agent-builder/agent-engine/memory-bank/overview) are now
[Generally Available](https://cloud.google.com/products#product-launch-stages).
Change **Vertex AI Agent Engine**

Vertex AI Agent Engine is now available in the following regions:

- `europe-west6` (Zurich)
- `europe-west8` (Milan)
- `asia-east2` (Hong Kong)
- `asia-northeast3` (Seoul)
- `asia-southeast2` (Jakarta)
- `northamerica-northeast2` (Toronto)
- `southamerica-east1` (São Paulo)

For more information, see [Vertex AI Agent Builder locations](https://docs.cloud.google.com/agent-builder/locations).
Change **Virtual Try-On**

Our Virtual Try-On model, `virtual-try-on-preview-08-04`, is
improved. Latency is significantly reduced and quality is improved for shoes,
body shape preservation, and product fidelity.

## December 12, 2025

Feature **Gemini 2.5 Flash with Gemini Live API Native Audio**

[Gemini 2.5 Flash with Gemini Live API Native Audio](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/2-5-flash-live-api) (`gemini-live-2.5-flash-native-audio`) is Generally Available (GA).
This model features cutting-edge native audio functionality for
[Gemini Live API](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/live-api), including enhanced voice quality and adaptability, Proactive Audio, and Affective Dialog.

## December 10, 2025

Feature **DeepSeek-V3.2** is available in Model Garden.
DeepSeek-V3.2 is a state-of-the-art large language model from
DeepSeek.
DeepSeek-V3.2 is available as a managed API in Model Garden. To learn more, see
[DeepSeek-V3.2](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/maas/deepseek/deepseek-v32).

## December 09, 2025

Feature The following models are available through Model Garden:

- [Ministral 3](https://console.cloud.google.com/vertex-ai/publishers/mistralai/model-garden/ministral-3)
- [Mistral Large 3](https://console.cloud.google.com/vertex-ai/publishers/mistralai/model-garden/mistral-large-3)

## December 08, 2025

Feature **Veo 3.1 video extension**

Veo 3.1 supports video extension in Preview.

For more information, see the following:

- [Extend Veo
  videos](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/extend-a-veo-video)
- [Veo video generation API](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo-video-generation)

## December 02, 2025

Feature The following models are available through Model Garden:

- [DeepSeek-V3.2](https://console.cloud.google.com/vertex-ai/publishers/deepseek-ai/model-garden/deepseek-v3-2;publisherModelVersion=deepseek-v3-2)
- [DeepSeek-V3.2-Speciale](https://console.cloud.google.com/vertex-ai/publishers/deepseek-ai/model-garden/deepseek-v3-2;publisherModelVersion=deepseek-v3-2-speciale)
Feature The Vertex AI Model Garden model co-hosting vLLM container is available to use with [this sample notebook](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/model_garden/model_garden_model_cohost.ipynb). You can use this container to serve multiple replicas of a model and serve multiple models with dynamic loading and unloading. This allows you to maximize resource utilization and serving efficiency, and flexibly adjust the models to serve.

## November 24, 2025

Feature **Anthropic's Claude Opus 4.5**

[Claude Opus 4.5](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/partner-models/claude/opus-4-5)
is available in Model Garden.

## November 17, 2025

Feature **Veo video generation**

Veo 3.1 is Generally Available, and introduces the following models:

- [Veo
  3.1](https://cloud.google.com/vertex-ai/generative-ai/docs/models/veo/3-1-generate#3.1-generate-001)
- [Veo 3.1
  Fast](https://cloud.google.com/vertex-ai/generative-ai/docs/models/veo/3-1-generate#3.1-fast-generate-001)

For more information, see the following:

- [Generate videos with Veo on Vertex AI](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/overview)
- [Generate Veo videos from text prompts](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/generate-videos-from-text)
- [Generate Veo videos from an image](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/generate-videos-from-an-image)
- [Generate Veo videos using first and last frames](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/generate-videos-from-first-and-last-frames)
- [Veo video generation API](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo-video-generation)
Announcement **LearnLM in Gemini**

The LearnLM model is no longer a separate offering or listing on AI Studio as
[LearnLM capabilities have been integrated into the latest Gemini models (starting with Gemini 2.5)](https://blog.google/outreach-initiatives/education/google-gemini-learnlm-update/).

Built in collaboration with experts in education,
[LearnLM](https://cloud.google.com/solutions/learnlm) represents our
capabilities fine-tuned for learning informed by rigorous research. These
advancements and improvements are available directly in Gemini, enhancing
educational experiences and applications.

Pre-existing learnlm-2.0-flash-experimental projects will not remain functional
past December 3, 2025 unless an alternative model is manually selected---we
encourage developers to switch to the latest Gemini models and optimize their
prompts by reviewing our
[LearnLM Partner Prompt Guide](https://services.google.com/fh/files/misc/learnlm_prompt_guide.pdf).

## November 13, 2025

Feature **Kimi K2 Thinking** is available in Model Garden. This model is
a thinking model that excels at complex problem-solving and deep reasoning.
Kimi K2 Thinking is available as a managed API in Model Garden. To learn more, see
[Kimi K2 Thinking](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/maas/kimi/kimi-k2-thinking).
Feature **Updated Prompt Caching for Anthropic Claude Models**

Prompt caching for Anthropic Claude models now supports a one-hour Time To Live (TTL).

For more information, see [Prompt caching](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/partner-models/claude/prompt-caching).

## November 11, 2025

Deprecated **Anthropic's Claude 3.7 Sonnet**

Anthropic's Claude 3.7 Sonnet is deprecated as of November 11, 2025 and will be
shut down on May 11, 2026. For more information, see
[Partner model deprecations](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/deprecations/partner-models#sonnet-3-7).

## November 07, 2025

Feature **Vertex AI Agent Engine**

The following features are now available in [Preview](https://cloud.google.com/products#product-launch-stages):

- Configure, manage, and view [observability
  features](https://docs.cloud.google.com/agent-builder/agent-engine/manage/overview#configure_telemetry_for_your_deployed_agent)
  such as sessions, traces, logs, and events for your agent in the Google Cloud console.

- [Use the
  playground](https://docs.cloud.google.com/agent-builder/agent-engine/use/overview#use-deployed-agents-console)
  to test and interact with your agent in the Google Cloud console.

- [Evaluate your agents](https://docs.cloud.google.com/agent-builder/agent-engine/evaluate) using the
  Gen AI evaluation service's GenAI Client in Vertex AI SDK.

- [Create and manage memory
  revisions](https://docs.cloud.google.com/agent-builder/agent-engine/memory-bank/revisions) for Memory Bank.

- Use Identity Access Management (IAM) to create an [agent
  identity](https://docs.cloud.google.com/agent-builder/agent-engine/agent-identity) to manage access and
  authentication when using agents on Vertex AI Agent Engine Runtime.

The following features are now available in [GA](https://cloud.google.com/products#product-launch-stages):

- [Express mode support](https://docs.cloud.google.com/agent-builder/agent-engine/overview#express-mode) for
  Vertex AI Agent Engine Runtime.

- Use the new free tier with Vertex AI Agent Engine Runtime. For more
  information, see
  [Pricing](https://cloud.google.com/vertex-ai/pricing#vertex-ai-agent-engine).

## November 04, 2025

Feature **MiniMax M2** is available in Model Garden. This model is
is built for end-to-end development workflows and has strong capabilities
in planning and executing complex tool-calling tasks. The model is
optimized to provide a balance of performance, cost, and inference speed.
MiniMax M2 is available as a managed API in Model Garden. To learn more, see
[MiniMax M2](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/maas/minimax/minimax-m2).

## October 23, 2025

Feature The following models are available through Model Garden:

- [DeepSeek-OCR](https://console.cloud.google.com/vertex-ai/publishers/deepseek-ai/model-garden/deepseek-ocr)
- [Qwen3-VL](https://console.cloud.google.com/vertex-ai/publishers/qwen/model-garden/qwen3-vl)
- [Earth AI](https://console.cloud.google.com/vertex-ai/model-garden/google/earth-ai)

## October 21, 2025

Security On September 23, 2025, we discovered a technical issue in
the Vertex AI API that resulted in a limited amount of responses
being misrouted between recipients for certain third-party models
when using streaming requests. This issue is now resolved.
Google models, e.g. Gemini, were not impacted.

Some internal proxies did not properly handle HTTP requests that
have an `Expect: 100-continue` header, resulting in
a desynchronization in a streaming response connection, where
a response intended for one request was instead delivered as
the response for a subsequent request.

For more information, see
[Security bulletins](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/security-bulletins#gcp-2025-059).

## October 16, 2025

Feature **Mistral's Codestral 2**

You can use Mistral's
[Codestral 2](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/partner-models/mistral/codestral-2)
in Model Garden.
Feature **vLLM TPU**

[vLLM TPU](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/open-models/vllm/use-vllm-tpu), a
highly-efficient serving framework for large language models (LLM) that's
optimized for [Cloud TPU](https://docs.cloud.google.com/tpu) hardware, is available through
Model Garden.

## October 15, 2025

Feature **Anthropic's Claude Haiku 4.5**

You can use Anthropic's
[Claude Haiku 4.5](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/partner-models/claude/haiku-4-5)
in Model Garden.
Feature **Veo video generation**

Veo 3.1 is available in Preview, and introduces the following models:

- [Veo
  3.1 preview](https://cloud.google.com/vertex-ai/generative-ai/docs/models/veo/3-1-generate-preview)
- [Veo 3.1
  Fast preview](https://cloud.google.com/vertex-ai/generative-ai/docs/models/veo/3-1-fast-generate-preview).

For more information, see the following:

- [Generate videos with Veo on Vertex AI](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/overview)
- [Generate Veo videos from text prompts](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/generate-videos-from-text)
- [Generate Veo videos from an image](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/generate-videos-from-an-image)
- [Generate Veo videos using first and last frames](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/generate-videos-from-first-and-last-frames)
- [Direct Veo using a reference image](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/use-reference-images-to-guide-video-generation)
- [Veo video generation API](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo-video-generation)
Feature **Veo video generation**

Veo 2 supports adding and removing objects from videos in Preview.

For more information about Veo 2, see [Veo 2
Preview](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/veo/2-0-generate-preview)

For more information about adding and removing objects, see the following:

- [Insert objects into Veo videos](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/insert-objects-into-videos)
- [Remove objects from Veo videos](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/remove-objects-from-videos)
- [Veo video generation API](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo-video-generation)

## October 14, 2025

Deprecated **Imagen subject and style fine-tuning**

Imagen subject model and style model tuning will be removed on
December 31, 2025. We recommend that you use [Gemini 2.5 Flash Image](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/2-5-flash-image), which supports most use cases that require
fine-tuning. For more information, see [Edit images with
Gemini](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/image-editing).
Deprecated **Imagen 4 preview models**

The following Imagen 4 preview models will be removed on
November 30, 2025:
`imagen-4.0-generate-preview-06-06`,
`imagen-4.0-ultra-generate-preview-06-06`, and
`imagen-4.0-fast-generate-preview-06-06`. To avoid service
disruption, migrate all workflows that use Imagen 4 preview models before
November 30, 2025, 2025, to the following Imagen 4 Generally
Available models: `imagen-4.0-generate-001`,
`imagen-4.0-ultra-generate-001`,
`imagen-4.0-fast-generate-001`.

## October 09, 2025

Change **Imagen**

Imagen's virtual try-on model, `virtual-try-on-preview-08-04`
was updated on September 30, 2025, to more accurately preserve the person's
body shape and preserve the garment's identity.

## October 07, 2025

Feature The following Qwen models are available in [Model Garden](https://console.cloud.google.com/vertex-ai/publishers/qwen/model-garden/qwen-image):

- Qwen-Image
- Qwen-Image-Edit
- Qwen-Image-Edit-2509
Announcement The [Computer Use tool](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/computer-use) (`gemini-2.5-computer-use-preview-10-2025`) is now available in Preview. The Computer Use model and tool lets you enable your applications to interact with and automate tasks in the browser. With the Computer Use model and tool, you can build agents that can:

- Automate repetitive data entry or form filling on websites.

- Navigate websites to gather information.

- Assist users by performing sequences of actions in web applications.

## October 06, 2025

Change **Updated pricing for Vertex AI Agent Engine** : Starting on **November 6, 2025**, Vertex AI Agent Engine Runtime will start charging for runtime usage for the following regions:

- `asia-southeast1` (Singapore)
- `australia-southeast2` (Melbourne)
- `europe-west2` (London)
- `europe-west3` (Frankfurt)
- `europe-west4` (Netherlands)

For more details, see [Pricing](https://cloud.google.com/vertex-ai/pricing#agent_engine) for Vertex AI Agent Engine.
Feature **Access Transparency for Vertex AI Agent Engine** : Access Transparency is now available for Vertex AI Agent Engine. For more information, see the overview for [Enterprise security](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/overview#enterprise_security).

## October 03, 2025

Feature **Prompt management**

Vertex AI offers tooling to help manage prompts and prompt versions. In addition to the prompt management capabilities in Vertex AI Studio, prompts can be stored and versioned using the Vertex AI SDK.

For more information, see the [Prompt management](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-reference/prompt-classes) API reference.

## October 02, 2025

Announcement **[Gemini 2.5 Flash Image](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/2-5-flash#image-generation)** (`gemini-2.5-flash-image`) is now generally available. This GA release adds support for aspect ratio controls, image-only response modality, regional endpoints, [support for batch predictions](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/batch-prediction-gemini), [image generation from multiple reference images](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/image-generation#image-generation), and [improved multi-turn image editing](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/image-editing#multi-turn-editing).

See [Gemini 2.5 Flash Image](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/2-5-flash#image) for more information.
Feature **Google Gen AI SDK in C# Preview**

[Preview](https://cloud.google.com/products#product-launch-stages): The Google Gen AI SDK is available in C#. See [googleapis/dotnet-genai](https://github.com/googleapis/dotnet-genai).

This release includes support for `GenerateContentAsync`, `GenerateContentStreamAsync`, `GenerateImagesAsync`, and three Live APIs, which includes `SendClientContentAsync`, `SendRealtimeInputAsync`, and `SendToolResponseAsync`.

## September 30, 2025

Feature [DeepSeek-V3.2-Exp](https://console.cloud.google.com/vertex-ai/publishers/deepseek-ai/model-garden/deepseek-v3-2) is available through Model Garden.

## September 25, 2025

Announcement New preview models for [**Gemini 2.5 Flash**](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/2-5-flash#2.5-flash-preview) and [**2.5 Flash-Lite**](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/2-5-flash-lite#2.5-flash-lite-preview) are now available. These models are available at the following versioned endpoints:

- `gemini-2.5-flash-preview-09-2025`
- `gemini-2.5-flash-lite-preview-09-2025`

## September 24, 2025

Deprecated Access to Gemini's 1.5 models has been discontinued. For more information, see our [Model versions](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions) page.

## September 23, 2025

Announcement **Gemini 2.5 Flash with Live API Native Audio Preview**

[**Gemini 2.5 Flash with Live API Native Audio**](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/release-notes/cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/2-5-flash) (`gemini-live-2.5-flash-preview-native-audio-09-2025`) is available in [Preview](https://cloud.google.com/products#product-launch-stages). A single, unified model processes audio input and generates audio output directly, eliminating separate text-to-speech/speech-to-text conversions. This results in-low latency, high-quality, and incredibly human-like conversations. New features and capabilities include:

- **Improved Barge-in:** Interrupt Gemini more naturally and reliably, even in loud and noisy environments.

- **Robust Function Calling:** We've improved the triggering rate, allowing Gemini to successfully execute the functions you define with greater precision.

- **Accurate Transcription:** The accuracy of audio-to-text transcription has been significantly enhanced.

- **Seamless Multilingual Support:** Speak to Gemini in multiple languages, and it will effortlessly switch between them without any pre-configuration. Language is no longer a barrier!

- **Enhanced Audio Quality:** Experience a dramatically improved audio quality that truly feels like speaking with a person.

- **Proactive Audio:** Define Gemini's expertise and set conditions for when it should respond. Gemini can act as a "silent listener," only chiming in when the conversation touches upon its designated area of expertise.

- **Affective Dialog:** Gemini can adapt and adjust its generated voice to match the emotional tone of the speaker, creating more empathetic and natural interactions.

[Watch our comprehensive demo](https://youtu.be/ZsB33Tr-P3c) to see these features in action, including seamless language switching, expert mode, emotionally aware responses, memory recall, and interactive screen sharing for engineering tasks -- all demonstrated directly within Vertex AI Studio without writing a single line of code!

## September 22, 2025

Feature [DeepSeek-V3.1-Terminus](https://console.cloud.google.com/vertex-ai/publishers/deepseek-ai/model-garden/deepseek-v3-1;publisherModelVersion=deepseek-v3-1-terminus) is available through Model Garden.

## September 18, 2025

Change **Grounding with Google Maps**

Grounding with Google Maps has implemented the following changes:

- Removed the following fields from the API response:
  - `grounding_chunk.maps.text`
  - `grounding_chunk.maps.place_answer_sources.review_snippets.author_attribution`
  - `grounding_chunk.maps.place_answer_sources.flag_content_uri`
  - `grounding_chunk.maps.place_answer_sources.review_snippets.flag_content_uri`
- The widget context token is only returned when the optional `widget_token_enable` input flag is set.

To learn more, see [Grounding with Google Maps](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/grounding/grounding-with-google-maps).

## September 15, 2025

Change **Imagen**

We improved Imagen's virtual try-on model, `virtual-try-on-preview-08-04`, so that it is better at preserving the person's body shape and preserving the garment product's identity.

## September 10, 2025

Breaking **Vertex AI Agent Engine**

In version `v1.112.0` of the Vertex AI SDK for Python, the `agent_engines` module has been refactored to a [client-based design](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/overview#migration). For information about updating your existing code to the new design, see the [Migration guide](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/deprecations/agent-engine-migration).
Feature **Vertex AI Agent Engine**

Agent Engine now supports the following features:

- Agent Engine **Code Execution** , now in Preview, lets your agent run code in an isolated sandbox environment. For more information, see [Code Execution](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/code-execution/overview).

- You can now develop, deploy, and use agents that support the **Agent-to-Agent (A2A) protocol** on Agent Engine. For more information, see [Develop an Agent2Agent agent](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/develop/a2a).

- Agent Engine now supports **bidirectional streaming** . For more information, see [Bidirectional streaming](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/bidirectional-streaming).

- The [Agent Engine page](https://console.cloud.google.com/vertex-ai/agents/agent-engines) in the Cloud Console UI now has a new **Memory Bank** tab for displaying and managing memories.

## September 09, 2025

Feature [EmbeddingGemma](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/embeddinggemma) and [DeepSeek-V3.1](https://console.cloud.google.com/vertex-ai/publishers/deepseek-ai/model-garden/deepseek-v3-1) models are available through Model Garden.
Feature [AI Singapore's SEA-LION V4 models](https://console.cloud.google.com/vertex-ai/publishers/aisingapore/model-garden/gemma-sea-lion-v4) are available through Model Garden. They are open models for Southeast Asian languages, built by leveraging Vertex Model Development Service for enhanced training efficiency and model accuracy.

## September 08, 2025

Feature **Veo video generation**

Veo 3 support for short-duration videos is [generally available](https://cloud.google.com/products#product-launch-stages). You can use Veo 3 to create 4, 6, or 8 second videos. For more information, see the following:

- [Generate videos with Veo on Vertex AI from text prompts](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/generate-videos-from-text)
- [Generate videos with Veo on Vertex AI from an image](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/generate-videos-from-an-image)
- [Veo on Vertex AI video generation API](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo-video-generation)

## September 03, 2025

Change **Vertex AI RAG Engine: Managed Database (Spanner)**

Customers will be charged for the use of a Google-managed Spanner instance that's provisioned in a Google tenant project, using standard Spanner SKUs.

For more information, see [Vertex AI RAG Engine billing](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/rag-engine/rag-engine-billing).

## August 26, 2025

Announcement **Gemini 2.5 Flash Image Preview**

[Gemini 2.5 Flash Image](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/2-5-flash#image-generation) (`gemini-2.5-flash-image-preview`) is available in [Preview](https://cloud.google.com/products#product-launch-stages). Gemini 2.5 Flash Image Preview supports additional image generation and editing features such as [image generation from multiple reference images](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/image-generation#image-generation) and [improved multi-turn image editing](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/image-editing#multi-turn-editing).
Feature **Vertex AI model tuning and Gen AI evaluation service**

Vertex AI model tuning now supports integration with the Gen AI evaluation service in Preview. You can automatically run evaluations on your tuned models and intermediate checkpoints. For more information, see [Create a tuning job](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini-use-supervised-tuning#create_a_text_model_supervised_tuning_job).

## August 21, 2025

Feature **Vertex AI Agent Engine**

Agent Engine now supports the following enterprise security features:

- You can now deploy your agents in a private VPC environment, configuring a Private Service Connect interface, to ensure data privacy and meet security and compliance requirements. For more information, see [Configure Private Service Connect interface](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/deploy#psc-i).

- You can now use your own [customer-managed encryption keys (CMEK)](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/manage/access#cmek) to protect data at rest.

- You can now specify [customized resource controls](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/deploy#customized-resource-controls), such as the minimum and maximum number of application instances, resource limits for each container, and concurrency for each container.

- As a part of Vertex AI Platform, Vertex AI Agent Engine now supports [HIPAA](https://cloud.google.com/security/compliance/hipaa) workloads.

For more information, see [Agent Engine overview](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/overview#enterprise_security).

## August 14, 2025

Announcement **Imagen**

Imagen 4 is Generally Available.

Imagen 4 introduces the following models:

- [Imagen 4 Generate](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/imagen/4-0-generate-001)
- [Imagen 4 Fast Generate](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/imagen/4-0-fast-generate-001)
- [Imagen 4 Ultra Generate](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/imagen/4-0-ultra-generate-001)

For more information, see [Generate images using text prompts](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/generate-images.md.txt) and [Image generation API](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-reference/imagen-api).
Feature [Gemma 3 270M](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/gemma3;publisherModelVersion=gemma-3-270m-it), [Wan 2.2](https://console.cloud.google.com/vertex-ai/publishers/wan-ai/model-garden/wan2-2) and [Wan 2.1](https://console.cloud.google.com/vertex-ai/publishers/wan-ai/model-garden/wan2-1) models are available through Model Garden.

## August 13, 2025

Feature [Qwen3 Coder](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/maas/qwen/qwen3-coder) and [Qwen3 235B](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/maas/qwen/qwen3-235b) are available as Model as a Service (MaaS) models in Model Garden.
Feature OpenAI's [gpt-oss-120b](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/maas/openai/gpt-oss-120b) and [gpt-oss-20b](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/maas/openai/gpt-oss-20b) are available as Model as a Service (MaaS) models in Model Garden.

## August 08, 2025

Feature **Gemini 2.5 Flash-Lite** and **Gemini 2.5 Pro** now support supervised fine-tuning. For more information, see [About supervised fine-tuning for Gemini models](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini-supervised-tuning).

## August 07, 2025

Feature **Vertex AI Agent Engine**

You can use your own [custom service account](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/set-up#custom-service-account) for agent identity to manage permissions and access according to your organization's security policies.
Feature **Model tuning**

You can now perform supervised fine-tuning on open models such as Llama 3.1. For more information, see [Tune an open model](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/open-model-tuning).
Feature **Vertex AI prompt optimizer**

The Vertex AI prompt optimizer is now [generally available](https://cloud.google.com/products?e=13802955&hl=en#product-launch-stages). For more information, see [Optimize prompts](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/learn/prompts/prompt-optimizer).

We now offer a [zero-shot prompt optimizer](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/learn/prompts/prompt-optimizer).

## August 06, 2025

Feature [OpenAI's gpt-oss](https://console.cloud.google.com/vertex-ai/publishers/openai/model-garden/gpt-oss) models are available through Model Garden.
Feature **Imagen**

Virtual try-on lets you generate virtual try-on images from an image of a
person and product photos that you provide, and is available in Preview. For more information, see [Generate Virtual Try-On Images](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/generate-virtual-try-on-images.md.txt) and [Virtual Try-On API](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-reference/virtual-try-on-api).

This release note is incorrect; see entry for [October 9, 2025](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/release-notes.md.txt#October_09_2025).

## July 29, 2025

Announcement **Veo video generation** Veo 3 and Veo 3 Fast are now generally available. For more information, see [Generate videos using text prompts](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/generate-videos-from-text).

## July 23, 2025

Change [Grounding with Google Maps](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/grounding/grounding-with-google-maps) is available in all regions (except for the EEA) as a Preview (Pre-GA) feature.

## July 22, 2025

Announcement **Gemini 2.5 Flash-Lite** is now generally available and accessible using the API and Vertex AI Studio. This GA release includes support for explicit caching and [batch prediction](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/batch-prediction-gemini), as well as expanded region support.

See [Gemini 2.5 Flash-Lite](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/2-5-flash-lite) for more information.

## July 17, 2025

Announcement **Veo 3** preview models now support upscaling for 1080p resolution using the new `resolution` parameter. For more information, see [Veo on Vertex AI](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo-video-generation#resolution).

## July 16, 2025

Feature Added [Gemma 3](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/model_garden/model_garden_axolotl_gemma3_finetuning.ipynb) fine-tuning notebook using Axolotl docker with support for 1b, 4b, 12b, and 27b variants.

## July 14, 2025

Feature [Multimodal MedGemma 27B IT](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/medgemma;publisherModelVersion=medgemma-27b-it), [MedSigLIP](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/medsiglip), and [T5Gemma](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/t5gemma) models are available through Model Garden.

## July 08, 2025

Feature **Vertex AI Agent Engine**

[Vertex AI Agent Engine Memory Bank](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/memory-bank/overview) is now available in Preview. Memory Bank lets you dynamically generate long-term memories based on users' conversations with your agent.

## July 03, 2025

Feature **Vertex AI Agent Garden**

[Vertex AI Agent Garden](https://console.cloud.google.com/vertex-ai/agents/agent-garden) now supports filtering by tags.

## June 27, 2025

Feature [Gemma 3n](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/gemma3n) models are now available through Model Garden.
Feature Multimodal datasets are now available in preview. For more information, see [Multimodal datasets](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/datasets).

## June 24, 2025

Deprecated Starting on June 24, 2025, Imagen versions 1 and 2, image captioning, and visual question answering are deprecated.

On September 24, 2025, the following features and models will be removed:

- image captioning
- visual question answering
- Imagen 1 model `imagegeneration@002`
- Imagen 2 models `imagegeneration@005` and `imagegeneration@006`

For more information, see [Migrate to Imagen 3](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/release-notes/vertex-ai/generative-ai/docs/image/migrate-to-imagen-3).

## June 23, 2025

Announcement **Veo 2** support for advanced video controls is Generally Available. In addition to a providing a first frame of a video, you can specify the last frame of a video or a video to extend in length. For more information, see [Veo on Vertex AI API](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo-video-generation).

## June 17, 2025

Change **Updated pricing for Gemini 2.5 Flash GA:** The price for Gemini 2.5 Flash in GA will be adjusted to reflect its quality and unified output token pricing. This includes lower prices for thinking output, higher prices for non-thinking output. These pricing changes will take effect on the new GA endpoint as shared above. Preview pricing will only continue on existing preview endpoints for 30 days post-GA on **July 15, 2025**.
Change **Provisioned Throughput (PT):** Once a model is GA, all new PT purchases will be for GA endpoints only. If you've purchased PT for a specific preview version, it will still work for that specific preview. However, you must [migrate the existing PT to the GA endpoint](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/provisioned-throughput/purchase-provisioned-throughput#change-order) or purchase new PT for the GA endpoint by **July 15, 2025**.
Announcement **Gemini 2.5 Flash-Lite** is now available as a preview offering in both the API and Vertex AI Studio.

See [Gemini 2.5 Flash-Lite](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/2-5-flash-lite) for more information.
Announcement **Live API** is now available as a private general availability offering in the API and Vertex AI Studio. Reach out to your Google account team representative to request access.

See [Live API](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/live-api) for more information.
Announcement **Gemini 2.5 Flash** and **Gemini 2.5 Pro** are now generally available and accessible using the API and Vertex AI Studio.

See [Gemini 2.5 Flash](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/2-5-flash) and [Gemini 2.5 Pro](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/2-5-pro) for more information.
Deprecated **Preview endpoint availability and removal:** All existing Gemini 2.5 Flash and Pro preview endpoints (listed below) will continue to be available with their current preview pricing until July 15, 2025. After this date, these preview endpoints will be shut down.

- `gemini-2.5-flash-preview-04-17`
- `gemini-2.5-flash-preview-05-20`
- `gemini-2.5-pro-preview-03-25`
- `gemini-2.5-pro-preview-05-06`
- `gemini-2.5-pro-preview-06-05`
Change **Updated preview endpoints:** Effective **June 19, 2025** , `gemini-2.5-flash-preview-04-17` endpoint will serve the Gemini 2.5 Flash model version released on 05-20, which has been promoted to GA. Similarly, the `gemini-2.5-pro-preview-05-06` and `03-25` endpoints will serve the Gemini 2.5 Pro model version released on 06-05, also promoted to GA. This update ensures continuity during your transition.

## June 16, 2025

Announcement The DeepSeek API service on Vertex AI is in [Preview](https://cloud.google.com/products#product-launch-stages). For more information, see the [DeepSeek model card](https://console.cloud.google.com/vertex-ai/publishers/deepseek-ai/model-garden/deepseek-r1-0528-maas) in Model Garden.

## June 11, 2025

Change **Imagen 4**'s public preview models are updated to the following:

- `imagen-4.0-generate-preview-06-06`
- `imagen-4.0-fast-generate-preview-06-06`
- `imagen-4.0-ultra-generate-preview-06-06`

For more information about each model, see [Preview Imagen models](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models#preview-imagen-models).

To avoid service interruption, migrate from `imagen-4.0-ultra-generate-exp-05-20` and `imagen-4.0-generate-preview-05-20` before 2025-07-07.

## June 09, 2025

Change **Gemini API**

The `logprobs` and `response_logprobs` parameters for the Gemini API are now [generally available](https://cloud.google.com/products#product-launch-stages). For more information, see [Generate content with Gemini API](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-reference/inference#logprobs).

## June 05, 2025

Feature **Gemini 2.5 Pro** 's public preview version has been updated to `gemini-2.5-pro-preview-06-05` and includes expanded support for thinking. This model version is available in the API and Vertex AI Studio.

See [Gemini 2.5 Pro](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/2-5-pro) for model details.

## June 03, 2025

Announcement In Model Garden, the following fine tuning features have been added:

- [Gemma 3](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/gemma3) UI fine-tuning using PEFT docker.
- [Qwen 2.5](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/model_garden/model_garden_pytorch_qwen2_5_finetuning.ipynb) fine-tuning notebook using PEFT docker.
- [Qwen 3](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/model_garden/model_garden_axolotl_qwen3_finetuning.ipynb) fine-tuning notebook using Axolotl docker.
- lm-evaluation-harness as an evaluation service in the [Llama 3.3](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/model_garden/model_garden_pytorch_llama3_3_finetuning.ipynb), [Llama 3.1](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/model_garden/model_garden_pytorch_llama3_1_finetuning.ipynb), [Gemma 3](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/model_garden/model_garden_gemma3_finetuning_on_vertex.ipynb) and [Gemma 2](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/model_garden/model_garden_gemma2_finetuning_on_vertex.ipynb) fine-tuning notebooks.
Announcement Model Garden now includes [DeepSeek-R1-0528](https://console.cloud.google.com/vertex-ai/publishers/deepseek-ai/model-garden/deepseek-r1) variants.

## May 23, 2025

Announcement Mistral OCR is an Optical Character Recognition API for document understanding. It is [GA](https://cloud.google.com/products#product-launch-stages) on Vertex AI. For more information, see the [Mistral OCR model card](https://console.cloud.google.com/vertex-ai/publishers/mistralai/model-garden/mistral-ocr-2505) in Model Garden.

## May 22, 2025

Announcement Anthropic's Claude Opus 4 and Claude Sonnet 4 are [GA](https://cloud.google.com/products#product-launch-stages) on Vertex AI and support [Provision Throughput](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/supported-models#partner-models). For more information, see the [Claude Opus 4](https://console.cloud.google.com/vertex-ai/publishers/anthropic/model-garden/claude-opus-4) or [Claude Sonnet 4](https://console.cloud.google.com/vertex-ai/publishers/anthropic/model-garden/claude-sonnet-4) model card in Model Garden.

## May 20, 2025

Feature **Vertex AI Agent Engine**

The following features are now available in Preview:

- [Console support for managing deployed agents](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/manage/overview)
- [Sessions support for Agent Development Kit agents](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/sessions/overview)
Feature **Audio-to-audio** support for [**Gemini 2.5 Flash with Live API**](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/2-5-flash#live-api-a2a) is now available as a private preview. Users must be allowlisted to use this new feature.

The model is available in the API and Vertex AI Studio.

See [Live API](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/live-api#a2a) for details.
Change **Gemini 2.5 Flash** 's public preview version has been updated to `gemini-2.5-flash-preview-5-20`.

See [Gemini 2.5 Flash](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/2-5-flash) for model details.

The model is available in the API and Vertex AI Studio.
Announcement **New stable text embeddings models** are now generally available:

- `gemini-embedding-001`
- `text-embedding-005`

For more information, see [Get text embeddings](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/embeddings/get-text-embeddings).
Feature **Thought summaries** are now available as an experimental feature for Gemini 2.5 Pro and 2.5 Flash.

For details, see [Thinking](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/thinking#thought-summaries).

The model is available in the API and Vertex AI Studio.
Feature **Imagen 4**

Imagen 4 offers two Preview models: [Imagen 4 Generate Preview 05-20](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/imagen/4-0-generate-preview-05-20), and [Imagen 4 Ultra Generate Experimental 05-20](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/imagen/4-0-ultra-generate-exp-05-20).

For more information, see
[Generate images using text prompts](http://cloud.google.com/vertex-ai/generative-ai/docs/image/generate-images) and the
[Generate images](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-reference/imagen-api) API.

The model is available in the API and Vertex AI Studio.
Announcement **Lyria 2**, our latest music generation model, is now generally available.

See our [music generation prompt guide](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/music/music-gen-prompt-guide) and our [user guide](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/music/generate-music) for more information.

The model is available in the API and Vertex AI Studio.
Announcement [MedGemma](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/medgemma) models are available in Model Garden.
Feature **Veo 3**

[Veo 3](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/veo/3-0-generate-preview) is available in Preview for allowlisted accounts.

For more information about Veo 3, see [Veo \| AI Video Generator](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/video/generate-videos) and [Veo on Vertex AI API](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-reference/veo-video-generation).

The model is available in the API and Vertex AI Studio.

## May 14, 2025

Deprecated [MedLM](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/medlm/overview.md.txt) is deprecated. Access to MedLM will no longer be available on or after September 29, 2025.

## May 07, 2025

Feature **Gemini 2.0 Flash with image generation** (`gemini-2.0-flash-preview-image-generation`) is now available as a public preview offering.

For more information, see [Generate images with Gemini](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/image-generation).
Fixed Seed parameter is now in GA and supports Gemini 2.5 model family.

## May 05, 2025

Change **Grounding**

The following grounding features are [generally available](https://cloud.google.com/products#product-launch-stages):

- [Grounding on your data using Vertex AI Search](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/grounding/grounding-with-your-data)
- [Grounding with Elasticsearch](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/grounding/grounding-with-elasticsearch)

## May 02, 2025

Announcement The global endpoint is generally available ([GA](https://cloud.google.com/products#product-launch-stages)). For details, see [Global endpoint](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/learn/locations#global-endpoint).

## April 30, 2025

Change
- Additional materials are available for deploying a model in Model Garden by using the [Python SDK, gcloud CLI, or API](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-garden/use-models#deploy_a_model), which are available in [Preview](https://cloud.google.com/products#product-launch-stages):
  - [Developer blog: Introducing the new Vertex AI Model Garden CLI and SDK](https://www.googlecloudcommunity.com/gc/Community-Blogs/Introducing-the-new-Vertex-AI-Model-Garden-CLI-and-SDK/ba-p/888386)
  - [Deploy open models by using the SDK tutorial notebook](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/model_garden/model_garden_deployment_tutorial.ipynb)
  - [Get started with Vertex AI Model Garden SDK notebook](https://github.com/GoogleCloudPlatform/generative-ai/blob/main/open-models/get_started_with_model_garden_sdk.ipynb)
Feature
- Llama 4 Maverick and Scout models are available in [Model Garden](https://console.cloud.google.com/vertex-ai/model-garden) with [Model-as-a-Service API Service](https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/llama-4-maverick-17b-128e-instruct-maas) and [self-hosted deployments](https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/llama4).
- [HiDream-I1](https://console.cloud.google.com/vertex-ai/publishers/hidream-ai/model-garden/hidream-i1), [Llama Guard 4](https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/llama-guard), [Llama Prompt Guard 2](https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/prompt-guard), and [Qwen3](https://console.cloud.google.com/vertex-ai/publishers/qwen/model-garden/qwen3) are available in [Model Garden](https://console.cloud.google.com/vertex-ai/model-garden).

## April 29, 2025

Announcement Gemini 1.5 Pro and Gemini 1.5 Flash models are not available in projects that have no prior usage of these models, including new projects. For details, see
[Model versions and lifecycle](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions).

## April 17, 2025

Announcement [Gemini 2.5 Flash](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/2-5-flash) with thinking and other well-rounded capabilities is now available in [Preview](https://cloud.google.com/products#product-launch-stages).

## April 10, 2025

Announcement Managed APIs for Llama 4 Maverick and Scout are in [Preview](https://cloud.google.com/products#product-launch-stages) on Vertex AI. For more information, see the [Llama 4 model card](https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/llama-4-maverick-17b-128e-instruct-maas).

## April 09, 2025

Feature **Vertex AI Agent Engine**

The following features are now available for Vertex AI Agent Engine in Preview:

- [Agent Development Kit integration](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/develop/adk)
- [Example Store](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/example-store/overview)
- [LlamaIndex Query Pipeline integration](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/develop/llama-index/query-pipeline)

The following features are now generally available for Vertex AI Agent Engine:

- [Agent monitoring](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/manage/monitoring)
Feature **Gemini Live API** is now available as a public preview offering and has been updated with the following features:

- Support for responses in 8 voices and 31 languages using Chirp 3
- Updated UI support in Vertex AI Studio
- Expanded conversation session window
- Ability to extend conversation sessions
- Support to share your current screen with Gemini during conversations
- Transcription support for audio in and audio out
- Support to change or update the system instructions mid-session

For more information, see [Gemini 2.0 Flash Live API](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/live-api).
Feature **Agent Development Kit** (ADK) is now available in [Preview](https://cloud.google.com/products#product-launch-stages). For more information, see [Agent Development Kit](https://google.github.io/adk-docs/).
Feature **Gemini 2.5 Pro** is now available as a public preview offering.

For more information, see [Gemini 2.5 Pro](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini/2-5-pro).
Feature **Agent Garden** is now available in [Preview](https://cloud.google.com/products#product-launch-stages). For more information, see [Vertex AI Agent Builder overview](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/agent-builder/overview) or go directly to [Agent Garden](https://console.cloud.google.com/vertex-ai/agents/agent-garden) in the Cloud Console.
Change **Vertex AI Agent Builder** now refers to a suite of features for building and deploying AI agents in Vertex AI. For more information see, [Vertex AI Agent Builder overview](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/agent-builder/overview).

The original Vertex AI Agent Builder product has been renamed **AI Applications** . The product functionality and endpoints remain the same. For more information, see [What is AI Applications?](https://docs.cloud.google.com/generative-ai-app-builder/docs/introduction).
Feature **Grounding: Web Grounding for Enterprise** is now Generally available. For more information, see [Web Grounding for Enterprise](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/grounding/web-grounding-enterprise).
Feature **Grounding: Grounding with Google Maps** is now available as a Public Experimental feature. For more information, see [Grounding with Google Maps](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/grounding/grounding-with-google-maps).

## March 25, 2025

Feature
- [DeepSeek-V3-0324](https://console.cloud.google.com/vertex-ai/publishers/deepseek-ai/model-garden/deepseek-v3), [TxGemma](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/txgemma) and [Sesame CSM](https://console.cloud.google.com/vertex-ai/publishers/sesame/model-garden/csm) are now available in [Model Garden](https://console.cloud.google.com/vertex-ai/model-garden).
- [DeepSeek-R1](https://console.cloud.google.com/vertex-ai/publishers/deepseek-ai/model-garden/deepseek-r1), [V3](https://console.cloud.google.com/vertex-ai/publishers/deepseek-ai/model-garden/deepseek-v3) and [V3-0324](https://console.cloud.google.com/vertex-ai/publishers/deepseek-ai/model-garden/deepseek-v3) can be deployed with H200 GPUs and improved vLLM support.
- You can deploy a model in Model Garden by using the [Python SDK, gcloud CLI, or API](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-garden/use-models#deploy_a_model), which are available in [Preview](https://cloud.google.com/products#product-launch-stages). You can get started with the "Equivalent code" in the deploy panel in the Model Garden console.

## March 20, 2025

Announcement Anthropic's Claude Sonnet 3.7 is [GA](https://cloud.google.com/products#product-launch-stages) on Vertex AI and supports [Provision Throughput](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/supported-models#partner-models). To learn more, view the [Claude Sonnet 3.7 model card](https://console.cloud.google.com/vertex-ai/publishers/anthropic/model-garden/claude-3-7-sonnet) in Model Garden.

## March 17, 2025

Announcement Mistral Small 3.1 (25.03) feature multimodal capabilities and a context of up to 128,000 tokens. For more information, see the [Mistral Small 3.1 (25.03) model card](https://console.cloud.google.com/vertex-ai/publishers/mistralai/model-garden/mistral-small-2503) in Model Garden.

## March 14, 2025

Feature [Judge model evaluation and customization tools](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/evaluate-judge-model) are now available in Preview for the [Gen AI evaluation service in Vertex AI](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/evaluation-overview).

## March 13, 2025

Announcement [Context caching](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/context-cache/context-cache-overview) for Gemini on Vertex AI is generally available ([GA](https://cloud.google.com/products#product-launch-stages)).

## March 12, 2025

Change Model Garden fine tuning updates:

- Added a [workbench-based notebook for Llama 3.1 finetuning](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/model_garden/model_garden_llama3_1_finetuning_with_workbench.ipynb).
- Updated [Llama 3.1](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/model_garden/model_garden_pytorch_llama3_1_finetuning.ipynb) and [Gemma 2](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/model_garden/model_garden_gemma2_finetuning_on_vertex.ipynb) UI fine-tuning with the updated PEFT docker.
Feature
- [Gemma 3](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/gemma3) and [ShieldGemma 2](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/shieldgemma2) are now available in Model Garden.
- [CogVideoX-2b](https://console.cloud.google.com/vertex-ai/publishers/thudm/model-garden/cogvideox) is now available in Model Garden.

## March 11, 2025

Change **Gemini 2.0 Flash Tuning**

Gemini 2.0 Flash fine-tuning is now generally available ([GA](https://cloud.google.com/products#product-launch-stages)).

Added support for [tuning function calling](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/tune-function-calling).

## March 04, 2025

Announcement **Vertex AI Agent Engine**

[Vertex AI Agent Engine](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/overview) is now [generally available (GA)](https://cloud.google.com/products#product-launch-stages).

Billing for Vertex AI Agent Engine starts on March 4, 2025. We recommend that you delete unused resources to avoid incurring unwanted costs. For more information, see [Pricing](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/overview#pricing).
Change LangChain on Vertex AI has been renamed to [Vertex AI Agent Engine](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/agent-engine/overview).

## February 25, 2025

Feature **Gemini 2.0 Flash-Lite is now generally available**

Gemini 2.0 Flash-Lite is now generally available. For more information, see [Gemini 2.0](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/gemini-v2).

## February 24, 2025

Announcement Anthropic's Claude Sonnet 3.7 is in [Preview](https://cloud.google.com/products#product-launch-stages) on Vertex AI. To learn more, view the [Claude Sonnet 3.7 model card](https://console.cloud.google.com/vertex-ai/publishers/anthropic/model-garden/claude-3-7-sonnet) in Model Garden.

## February 21, 2025

Change
- PEFT Docker updates
  - Added support for evaluation metrics like perplexity, bleu, google_bleu, rouge1, rouge2, rougeL, rougeLSum.
  - Uses the best checkpoint and loads the model based on the best eval metrics.
  - Run training and eval only for data which is less than or equal to the `max_seq_length`.
  - Use `gcloud storage rsync` instead of `csfuse` to save a checkpoint.
- Fine tuning updates
  - You can select a service account when you click **Fine-tune** for a model, such as [Llama 3.1](https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/llama3_1).
  - Added a [PEFT based LLM finetuning tutorial notebook](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/model_garden/model_garden_finetuning_tutorial.ipynb).
  - Added a [Axolotl based LLM finetuning notebook](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/model_garden/model_garden_axolotl_finetuning.ipynb).
  - Updated [Llama 3.1](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/model_garden/model_garden_pytorch_llama3_1_finetuning.ipynb) and [Gemma 2](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/model_garden/model_garden_gemma2_finetuning_on_vertex.ipynb) fine-tuning notebooks with the updated PEFT Docker container.
- Model updates
  - Updated the [PaliGemma](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/paligemma) model card by supporting PaliGemma 2 mix models, and segmentation functionality to Paligemma 1 models.
  - Updated the [LLaVa model card](https://console.cloud.google.com/vertex-ai/publishers/liuhaotian/model-garden/pytorch-llava) by supporting LLaVA Next models and adding vLLM to the notebook.

## February 12, 2025

Feature Deepseek-V3 and Deepseek-R1 have been added to [Model Garden](https://console.cloud.google.com/vertex-ai/model-garden) in [Preview](https://cloud.google.com/products#product-launch-stages):

- [DeepSeek-V3](https://console.cloud.google.com/vertex-ai/publishers/deepseek-ai/model-garden/deepseek-v3) (671B) is a powerful Mixture-of-Experts (MoE) language model with 671B total parameters with 37B activated for each token.
- [DeepSeek-R1](https://console.cloud.google.com/vertex-ai/publishers/deepseek-ai/model-garden/deepseek-r1) (671B) is one of the first-generation reasoning models introduced by DeepSeek and offers performance comparable to OpenAI-o1 across math, code, and reasoning tasks.

You can use a [notebook](https://colab.research.google.com/github/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/model_garden/model_garden_pytorch_deepseek_deployment.ipynb) to deploy these models.

## February 11, 2025

Announcement The [Llama 3.3 70B](https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/llama-3.3-70b-instruct-maas) model that is managed on Vertex AI is now in [Preview](https://cloud.google.com/products#product-launch-stages).

## February 07, 2025

Feature [deepseek-ai/deepseek-r1](https://console.cloud.google.com/vertex-ai/publishers/deepseek-ai/model-garden/deepseek-r1) and [microsoft/Phi-4](https://console.cloud.google.com/vertex-ai/publishers/microsoft/model-garden/phi4) models were added to [Model Garden](https://console.cloud.google.com/vertex-ai/model-garden).
Change The following advanced LLM inference optimization techniques are available in Model Garden in [Preview](https://cloud.google.com/products#product-launch-stages):

- Prefix caching reuses computations from previously generated text, eliminating redundant processing. It reduces time-to-first-token for requests with common prompt prefixes. Prefix caching is available for the following models:

  - vLLM: Llama 3.1 (8b, 70b), Llama 3.3 (70b)
  - Hex-LLM: Llama 2 (7b, 13b), Llama 3 (8b), Llama 3.1 (8b, 70b), Llama 3.2 (1b, 3b), Llama Guard (1b, 8b), CodeLlama (7b, 13b), Gemma (2b, 7b), CodeGemma (2b, 7b), Mistral-7B (v0.2, v0.3), Mixtral-8x7B (v0.1)
- Speculative decoding is an effective optimization technique to reduce generation time-per-output-token latency. For more information, see the [Model Garden advanced features notebook](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/model_garden/model_garden_advanced_features.ipynb).

## February 05, 2025

Feature **Gemini 2.0 Flash general availability for text-only output**

Gemini 2.0 Flash is now generally available for text-only outputs. Multimodal outputs are still available only as a private preview. For more information, see [Gemini 2.0](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/gemini-v2).
Feature **New Gemini 2.0 Pro and Gemini 2.0 Flash-Lite models available to users**

Two new models in the Gemini 2.0 family are now available to users:

- **Gemini 2.0 Pro:** Our strongest model for coding and world knowledge, featuring a 2M long context window. Gemini 2.0 Pro is available as an experimental model in Vertex AI.
- **Gemini 2.0 Flash-Lite:** Our fastest and most cost efficient Flash model. Gemini 2.0 Flash-Lite is available as a Preview model in Vertex AI.

For more information, see [Gemini 2.0](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/gemini-v2)

## January 31, 2025

Feature You can now monitor usage, throughput, and latency and troubleshoot 429 errors on Vertex AI foundation models, like Google Gemini and Anthropic Claude, by using a predefined dashboard. After querying a model from the [Vertex AI Model Garden](https://cloud.google.com/model-garden), you can find the name of the model you queried in the Vertex AI **Dashboard** page under the "Model observability" heading.

To customize the dashboard and explore relevant metrics in Cloud Monitoring, click **Show All Metrics** . For information about using dashboards in Cloud Monitoring, see [View and customize Google Cloud dashboards](https://docs.cloud.google.com/monitoring/charts/predefined-dashboards.md.txt).

## January 30, 2025

Deprecated Mistral Large (24.07) and Codestral (24.05) that are offered as a Model as a Service (MaaS) models in Model Garden are deprecated. For details, see [Generative AI on Vertex AI deprecations](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/deprecations.md.txt).

## January 29, 2025

Feature **New Imagen 3 image generation model available to users**

A newer improved Imagen 3 image generation model is now available to all users:

- `imagen-3.0-generate-002`

This image generation model supports the following additional features:

- **Prompt enhancement** - The LLM-based prompt rewriter tool adds additional details and descriptive language to the prompt you provide, generally resulting in higher quality generated images. This feature is configurable and is enabled by default.

For more information, see [Imagen on Vertex AI model versions and lifecycle](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/model-versioning) and [Generate images using text prompts](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/generate-images.md.txt).

## January 22, 2025

Announcement **LangChain on Vertex AI**

Billing for LangChain on Vertex AI will start on March 4, 2025.

The pricing structure is based on vCPU hours and GiB hours used. This means that you will be charged for both the compute (vCPU) and memory resources consumed by your LangChain on Vertex AI workloads.

You can review the pricing details in the table below.

| Product | SKU ID | Price |
|---|---|---|
| ReasoningEngine vCPU | 8A55-0B95-B7DC | $0.0994/vCPU-Hr |
| ReasoningEngine Memory | 0B45-6103-6EC1 | $0.0105/GiB-Hr |

## January 21, 2025

Deprecated Anthropic's Claude 3 Sonnet that is offered as a Model as a Service (MaaS) model in Model Garden is deprecated. For details, see [Generative AI on Vertex AI deprecations](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/deprecations.md.txt).

## January 17, 2025

Feature [Agent evaluation](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/evaluation-agents) using the [Gen AI evaluation service](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/evaluation-overview) is available in [Preview](https://cloud.google.com/products#product-launch-stages).

## December 20, 2024

Change [RAG Engine](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/rag-overview) is [generally available (GA)](https://cloud.google.com/products#product-launch-stages).

The supported models include the following:

- Google Gemini
- Google embedding and OSS E5 embedding models
- Model Garden self-deployed OSS LLMs
- [Model as a service (MaaS) Llama models](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/partner-models/use-partner-models)

The supported features include the following:

- **[Data connectors](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/use-data-connectors)**: Google Cloud Storage, Google Drive, Slack, Jira, and SharePoint
- **[Document types](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/supported-documents)**: Google Workspace documents, HTML, JSON, Markdown, PDF, and text files
- **Transformations**: fixed-size chunking and chunk overlap
- **Vector databases**: Vertex AI Vector Search and Pinecone

## December 18, 2024

Feature [Hex-LLM: High-Efficiency Large Language Model Serving](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/open-models/use-hex-llm) is available in [General Availability (GA)](https://cloud.google.com/products#product-launch-stages).

This launch adds support for the following models:

- Llama 3.1
- Llama 3.2
- Phi-3
- Qwen2 and Qwen2.5

Additional supported features:

- Multi-host serving.
- Disaggregated serving (experimental).
- Prefix caching.
- AWQ quantization.

## December 17, 2024

Feature You can copy tuned Gemini 1.5 Pro 002 and Gemini 1.5 Flash 002 adapter models across projects. For details, see [Copy a model in Vertex AI Model Registry](https://docs.cloud.google.com/vertex-ai/docs/model-registry/copy-model).

## December 11, 2024

Feature The Gemini 2.0 Flash (`gemini-2.0-flash-exp`) model is Generally available for grounded answer generation with RAG. This model is tuned to address context-based question and answering tasks. For more information, see [Ground responses for Gemini models](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/ground-gemini).

## December 10, 2024

Feature **Imagen 3 image generation models Generally Available to all users**

Imagen 3 image generation models are now available to all users without requiring prior approval. These include the following image generation models:

- `imagen-3.0-generate-001`
- `imagen-3.0-fast-generate-001` (low latency model)

Prior image generation models (`imagegeneration@006`, `imagegeneration@005`, `imagegeneration@002`) still require approval to use.

For more information, see [Imagen on Vertex AI model versions and lifecycle](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/model-versioning) and [Generate images using text prompts](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/generate-images.md.txt).
Feature **Imagen 3 Customization model Generally Available to approved users**

Imagen 3 Customization model is now available to approved users. This includes the following model:

- `imagen-3.0-capability`

Imagen 3 Customization lets you guide image generation by providing reference images (few-shot learning). Imagen 3 Customization lets you customize generated images for the following feature categories:

- [Subject Customization](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/subject-customization.md.txt) (product, person, and animal companion)
- [Style Customization](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/style-customization.md.txt)
- [Controlled Customization](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/edit-controlled.md.txt) (canny edge and scribble)
- [Instruct Customization](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/instruct-customization.md.txt) (Style transfer)
Feature **Imagen 3 editing model Generally Available to approved users**

The Imagen 3 Editing model is now available to approved users. This includes the following model:

- `imagen-3.0-capability`

This model offers the following additional features:

- [Inpainting](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/edit-inpainting) - Add or remove content from a masked area of an image
- [Outpainting](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/edit-outpainting.md.txt) - Expand a masked area of an image
- [Product image editing](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/edit-product-image) - Identify and maintain a primary product while changing the background or product position

For more information, see [Model versions](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/model-versioning).

## December 06, 2024

Security A vulnerability was discovered in the Vertex AI API serving Gemini multimodal requests, allowing bypass of VPC Service Controls. For details, see the [Security bulletins](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/security-bulletins#gcp-2024-063) page.

## November 21, 2024

Feature The Gen AI evaluation service can now help you evaluate your translation models using MetricX, COMET, and BLEU metrics.
To learn more about evaluating your translation models, see [Evaluate translation models](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/determine-eval#translation-models).
Announcement Mistral Large (24.11) is Generally Available on Vertex AI as a managed model. To learn more, view the [Mistral Large (24.11) model card](https://console.cloud.google.com/vertex-ai/publishers/mistralai/model-garden/mistral-large-2411) in Model Garden.

## November 08, 2024

Feature [Batch predictions for Llama models](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/partner-models/llama-batch) on Vertex AI (MaaS) is available in [Preview](https://cloud.google.com/products#product-launch-stages).
Feature **Batch prediction support for Gemini**

Batch prediction is available for Gemini in [General Availability (GA)](https://cloud.google.com/products#product-launch-stages). Available Gemini models include Gemini 1.0 Pro, Gemini 1.5 Pro, and Gemini 1.5 Flash. To get started with batch prediction, see [Get batch predictions for Gemini](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/batch-prediction-gemini).

## November 05, 2024

Change We are extending the availability of Gemini 1.0 Pro 001 and Gemini 1.0 Pro Vision 001 from February 15, 2025 to April 9, 2025. For details, see the [Deprecations](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/deprecations.md.txt).

## November 04, 2024

Announcement The Anthropic Claude Haiku 3.5 is Generally Available on Vertex AI. To learn more, view the [Claude Haiku 3.5 model card](https://console.cloud.google.com/vertex-ai/publishers/anthropic/model-garden/claude-3-5-haiku) in Model Garden.
Change The translation LLM now supports Polish, Turkish, Indonesian, Dutch, Vietnamese, Thai and Czech. For the full list of supported languages, see the [Translate text](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/translate/translate-text) page.

## October 28, 2024

Feature The [Whisper large v3 and Whisper large v3 turbo](https://console.cloud.google.com/vertex-ai/publishers/openai/model-garden/whisper-large) models have been added to Model Garden.
Feature You can now fine-tune the following models from the Cloud console:

- [Gemma 2](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/gemma2)
- [Llama 3.1](https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/llama3_1)
- [Mistral](https://console.cloud.google.com/vertex-ai/publishers/mistral-ai/model-garden/mistral)
- [Mixtral](https://console.cloud.google.com/vertex-ai/publishers/mistral-ai/model-garden/mixtral)
Change Updated the fine-tuning notebooks for [Gemma 2](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/gemma2), [Llama 3.1](https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/llama3_1), [Mistral](https://console.cloud.google.com/vertex-ai/publishers/mistral-ai/model-garden/mistral), and [Mixtral](https://console.cloud.google.com/vertex-ai/publishers/mistral-ai/model-garden/mixtral) with the following enhancements:

- The notebooks use an updated high-performance container for single host multi-GPU LoRA fine-tuning.
  - Better throughput and GPU utilization with well-tested max-sequence-lengths.
  - Support for input token masking.
  - No out of memory (OOM) error during fine-tuning.
- Added a custom dataset example that uses a template and format validation.
- Support for a default accelerator pool with quota checks.
- Improved documentation.

## October 22, 2024

Announcement The Anthropic Claude Sonnet 3.5 v2 is Generally Available. To learn more, view the [Claude Sonnet 3.5 v2 model card](https://console.cloud.google.com/vertex-ai/publishers/anthropic/model-garden/claude-3-5-sonnet-v2) in Model Garden.

## October 18, 2024

Announcement The [Llama 3.1 405B](https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/llama-3.1-405b-instruct-maas) model that is managed on Vertex AI is now [Generally Available](https://cloud.google.com/products#product-launch-stages).

## October 09, 2024

Feature The Vertex AI Gemini API SDK supports tokenization capabilities for local token counting and computation. This is a streamlined way to compute tokens locally, ensuring compatibility across different Gemini models and their tokenizers. Supported models include ` gemini-1.5-flash` and ` gemini-1.5-pro` . To learn more, see [Count tokens](https://cloud.google.com/vertex-ai/generative-ai/docs/multimodal/list-token).

## October 04, 2024

Change Added dynamic LoRA serving for [Llama 3.1](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/model_garden/model_garden_pytorch_llama3_1_deployment.ipynb) and [Stable Diffusion XL](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/model_garden/model_garden_pytorch_stable_diffusion_xl_lora.ipynb).
Feature You can deploy Hugging Face models on Google Cloud that have [text embedding inference](https://huggingface.github.io/text-embeddings-inference/) enabled or [pytorch inference](https://huggingface.co/docs/inference-endpoints/supported_tasks)
enabled. For more information, see the [Hugging Face model deployment](http://console.cloud.google.com/vertex-ai/model-garden;action=deploy;hfSource=true) in the console.
Feature [Prompt Guard](https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/prompt-guard) and [Flux](https://console.cloud.google.com/vertex-ai/publishers/black-forest-labs/model-garden/flux1-schnell) were added to [Model Garden](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-garden/explore-models).
Change Added multiple deployment settings (with A100-80G and H100) and sample requests for some popular models, including [Llama 3.1](https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/llama3_1), [Gemma 2](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/gemma2), and [Mixtral](https://console.cloud.google.com/vertex-ai/publishers/mistral-ai/model-garden/mixtral).
Feature The AI assistant in Vertex AI Studio can help you refine and generate prompts. This feature is in [Preview](https://cloud.google.com/products#product-launch-stages). To learn more, see [Use AI-powered prompt writing tools](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/learn/prompts/ai-powered-prompt-writing).

## October 01, 2024

Feature **Grounding: Dynamic retrieval for grounded results (GA)**

Dynamic retrieval lets you choose when to turn off grounding with Google Search. This is useful when a prompt doesn't require an answer grounded in Google Search, and the supported models can provide an answer based on their knowledge without grounding. Dynamic retrieval helps you manage latency, quality, and cost more effectively.

This feature is [Generally Available](https://cloud.google.com/products#product-launch-stages). For more information, see [Dynamic retrieval](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/ground-gemini#dynamic-retrieval).

## September 30, 2024

Feature Prompt templates let you to test how different prompt formats perform with different sets of prompt data. This feature is in [Preview](https://cloud.google.com/products#product-launch-stages). To learn more, see [Use prompt templates](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/learn/prompts/prompt-templates).

## September 25, 2024

Announcement The Llama 3.2 90B model is available in [Preview](https://cloud.google.com/products#product-launch-stages) on Vertex AI. Llama 3.2 90B enables developers to build and deploy the latest generative AI models and applications that use Llama's capabilities, such as image reasoning. Llama 3.2 is also designed to be more accessible for on-device applications. For more information, see [Llama models](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/partner-models/llama).

## September 24, 2024

Announcement New stable versions of Gemini 1.5 Pro (`gemini-1.5-pro-002`) and Gemini 1.5 Flash (`gemini-1.5-flash-002`)
are [Generally Available](https://cloud.google.com/products#product-launch-stages). These models introduce broad quality improvements over the previous `001` versions, with significant gains in the following categories:

- Factuality and reduce model hallucinations
- Openbook Q\&A for RAG use cases
- Instruction following
- Multilingual understanding in 102 languages, especially in Korean, French, German, Spanish, Japanese, Russian, and Chinese.
- SQL generation
- Audio understanding
- Document understanding
- Long context
- Math and reasoning

For more information about differences with the previous model versions, see [Model versions and lifecycle](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versions).
Feature Use Gemini to directly analyze YouTube videos and publicly available media (such as images, audio, and video) by using a link. This feature is in [Public Preview](https://cloud.google.com/products#product-launch-stages).
Announcement The 2M context window with Gemini 1.5 Pro is now in [Generally Available](https://cloud.google.com/products#product-launch-stages), which opens up long-form multimodal use cases that only Gemini can support.
Feature The new API parameters `audioTimestamp`, `responseLogprob`, and `logprobs` are in [Public Preview](https://cloud.google.com/products#product-launch-stages). For more information, see [API reference](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-reference/inference).
Feature The Vertex AI prompt optimizer adapts your prompts using the optimal instructions and examples to elicit the best performance from your chosen model. This feature is available in [Preview](https://cloud.google.com/products#product-launch-stages). To learn more, see [Optimize prompts](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/learn/prompts/prompt-optimizer).
Change Gemini 1.5 Pro and Gemini 1.5 Flash Tuning is now available in [GA](https://cloud.google.com/products#product-launch-stages).
Tune Gemini with text, image, audio, and document data types using the latest models:

- `gemini-1.5-pro-002`
- `gemini-1.5-flash-002`

Gemini 1.0 tuning remains in preview.

For more information on tuning Gemini, see [Tune Gemini models by using supervised fine-tuning](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/gemini-use-supervised-tuning).
Feature Gemini 1.5 Pro and Gemini 1.5 Flash now support multimodal input with [function calling](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/function-calling). This feature is in [Preview](https://cloud.google.com/products#product-launch-stages).
Change The latest versions of **Gemini 1.5 Flash** (`gemini-1.5-flash-002`) and **Gemini 1.5 Pro** (`gemini-1.5-pro-002`) use [dynamic shared quota](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/dsq), which distributes on-demand capacity among all queries being processed. Dynamic shared quota is [Generally Available](https://cloud.google.com/products#product-launch-stages).
Announcement [Controlled generation](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/control-generated-output) is now [Generally Available](https://cloud.google.com/products#product-launch-stages).

## September 20, 2024

Feature Add label metadata to `generateContent` and `streamGenerateContent` API calls. For details, see [Add labels to API calls](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/add-labels-to-api-calls).

## September 18, 2024

Announcement Model Garden supports an organization policy so that administrators can limit access to certain models and capabilities. For more information, see [Control access to Model Garden models](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/control-model-access)

## September 03, 2024

Change Gemini 1.5 Flash (`gemini-1.5-flash`) supports [controlled generation](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/control-generated-output).

## August 30, 2024

Feature Gen AI Evaluation Service is Generally Available. To learn more, see the [Gen AI Evaluation Service overview](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/evaluation-overview).

## August 26, 2024

Change For controlled generation, you can have the model respond with an enum value in plain text, as defined in your response schema. Set the `responseMimeType` to `text/x.enum`. For more information, see [Control generated output](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/control-generated-output).

## August 22, 2024

Change **AI21 Labs**

Managed models from AI21 Labs are available on Vertex AI. To use a AI21 Labs model on Vertex AI, send a request directly to the Vertex AI API endpoint. For more information, see [AI21 models](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/partner-models/ai21).

## August 09, 2024

Announcement Gemini on Vertex AI supports multiple response candidates. For details, see [Generate content with the Gemini API](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-reference/inference).

## August 05, 2024

Change The translation LLM now supports Arabic, Hindi, and Russian. For the full list of supported languages, see the [Translate text](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/translate/translate-text) page.

## August 02, 2024

Feature Vertex AI SDK for Python supports token listing and counting for prompts without the need to make API calls. This feature is available in ([Preview](https://cloud.google.com/products#product-launch-stages)). For details, see [List and count tokens](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/list-token).

## July 31, 2024

Feature **New Imagen on Vertex AI image generation model and features**

The Imagen 3 image generation models (`imagen-3.0-generate-001` and the low-latency version `imagen-3.0-fast-generate-001`) are Generally Available to [approved users](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/overview.md.txt#feature-launch-stage). These models offer the following additional features:

- Additional aspect ratios (1:1, 3:4, 4:3, 9:16, 16:9)
- Digital watermark ([SynthID](https://deepmind.google/technologies/synthid/)) enabled by default
- Watermark verification
- User-configurable safety features (safety setting, person/face setting)

For more information, see [Model versions](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/model-versioning) and [Generate images using text prompts](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/generate-images.md.txt).
Announcement Gemma 2 2B is available in Model Garden. For details, see [Use Gemma open models](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/open-models/use-gemma).
Feature The following models have been added to [Model Garden](https://console.cloud.google.com/vertex-ai/model-garden):

- [Gemma 2 2B](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/gemma2): A foundation LLM by Google DeepMind.
- [Qwen2](https://console.cloud.google.com/vertex-ai/publishers/qwen/model-garden/qwen2): An LLM series by Alibaba Cloud.
- [Phi-3](https://console.cloud.google.com/vertex-ai/publishers/microsoft/model-garden/phi3): An LLM series by Microsoft.
Change Resource and deployment settings were made to the following models:

- Added GPU inferences for [gemma2-27b](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/gemma2) and [gemma2-27b-it](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/gemma2) with verified performances.
- Added verified deployment settings for Mistral AI models that are deployed from Huggingface, including [mistralai/mistral-nemo-instruct-2407](https://console.cloud.google.com/vertex-ai/publishers/mistralai/model-garden/mistral-nemo-instruct-2407;action=deploy;hfSource=true), [mistralai/mistral-nemo-base-2407](https://console.cloud.google.com/vertex-ai/publishers/mistralai/model-garden/mistral-nemo-base-2407;action=deploy;hfSource=true), [mistralai/mistral-large-instruct-2407](https://console.cloud.google.com/vertex-ai/publishers/mistralai/model-garden/mistral-large-instruct-2407;action=deploy;hfSource=true), and [mistralai/codestral-22b-v0.1](https://console.cloud.google.com/vertex-ai/publishers/mistralai/model-garden/codestral-22b-v0.1;action=deploy;hfSource=true).
- Added multiple deployment settings with A100 (40G), A100 (80G) and H100 (80G) for select models, such as [llama3.1](https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/llama3_1), [llama3](https://console.cloud.google.com/vertex-ai/publishers/meta/model-garden/llama3), [gemma2](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/gemma2), [gemma](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/gemma), and [mistral-7b](https://console.cloud.google.com/vertex-ai/publishers/mistral-ai/model-garden/mistral-7b).

## July 30, 2024

Announcement See the [Gemini Online Inference on Vertex AI Service Level Agreement (SLA)](https://cloud.google.com/vertex-ai/generative-ai/sla).

## July 24, 2024

Announcement **Mistral AI**

Managed models from Mistral AI are available on Vertex AI. To use a Mistral AI model on Vertex AI, send a request directly to the Vertex AI API endpoint. For more information, see [Mistral AI models](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/partner-models/mistral).

## July 23, 2024

Announcement **Llama 3.1**

The Llama 3.1 405B model is available in [Preview](https://cloud.google.com/products#product-launch-stages) on Vertex AI. Llama 3.1 405B provides capabilities from synthetic data generation to model distillation, steerability, math, tool use, multilingual translation, and more. For more information, see [Llama models](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/partner-models/llama).

## July 02, 2024

Announcement Google's open weight Gemma 2 model is available in Model Garden. For details, see [Use Gemma open models](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/open-models/use-gemma).
Change MaMMUT is now available in [Model Garden](https://console.cloud.google.com/vertex-ai/model-garden). MaMMUT is a vision-encoder and text-decoder model for multimodal tasks such as visual question answering, image-text retrieval, text-image retrieval, and generation of multimodal embeddings.

## June 28, 2024

Feature Launched [Hex-LLM](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/open-models/use-hex-llm) for high-efficiency large language model serving. This performant TPU serving solution is based on XLA and optimized kernels to achieve high throughput and low latency.

Hex-LLM uses several parallelism strategies for multiple TPU chips, quantizations, dynamic LoRA, and more. Hex-LLM supports the following dense and sparse LLMs:

- Gemma 2B and 7B
- Gemma 2 9B and 27B
- Llama 2 7B, 13B and 70B
- Llama 3 8B and 70B
- Mistral 7B and Mixtral 8x7B
Feature The following models have been added to [Model Garden](https://console.cloud.google.com/vertex-ai/model-garden):

- 36 Hugging Face embedding models with verified deployment settings such as [BAAI/bge-m3](https://console.cloud.google.com/vertex-ai/publishers/BAAI/model-garden/bge-m3;action=deploy;hfSource=true) and [intfloat/multilingual-e5-large-instruct](https://console.cloud.google.com/vertex-ai/publishers/intfloat/model-garden/multilingual-e5-large-instruct;action=deploy;hfSource=true).
- 35 Hugging Face PyTorch models with verified deployment settings such as [stabilityai/stable-diffusion-2-1](https://console.cloud.google.com/vertex-ai/publishers/stabilityai/model-garden/stable-diffusion-2-1;action=deploy;hfSource=true).

For more information, see the [Hugging Face model deployment](http://console.cloud.google.com/vertex-ai/model-garden;action=deploy;hfSource=true) in the console.
Change
- Updated Docker images in [Llama 3 notebooks](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/model_garden/model_garden_pytorch_llama3_finetuning.ipynb) that are more efficient at tuning.
- A notebook-based interactive workshop UI was added in [Model Garden](https://console.cloud.google.com/vertex-ai/model-garden) for image generative models such as [stable-diffusion-xl-base](https://console.cloud.google.com/vertex-ai/publishers/stability-ai/model-garden/stable-diffusion-xl-base), [image inpainting](https://console.cloud.google.com/vertex-ai/publishers/runwayml/model-garden/stable-diffusion-inpainting), [controlnet](https://console.cloud.google.com/vertex-ai/publishers/lllyasviel/model-garden/control-net). You can find these models from the **Open Notebook** list.
- Colab Notebooks for frequently used models in Model Garden have been revised with no-code or low-code implementations to improve accessibility and user experience.

## June 27, 2024

Feature Context caching is available for Gemini 1.5 Pro. Use context caching to reduce the cost of requests that contain repeat content with high input token counts. For more information, see [Context caching overview](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/context-cache/context-cache-overview).

## June 25, 2024

Feature Controlled generation is available on Gemini 1.5 Pro and supports the JSON schema. For more information, see [Control generated output](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/control-generated-output).

## June 20, 2024

Announcement The Anthropic Claude Sonnet 3.5 is Generally Available. To learn more, view the [Claude Sonnet 3.5 model card](https://console.cloud.google.com/vertex-ai/publishers/anthropic/model-garden/claude-3-5-sonnet) in Model Garden.

## June 17, 2024

Change Increased the input token limit for Gemini 1.5 Pro from 1M to 2M. For more information, see [Google models](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/learn/models).

## June 11, 2024

Change **Upload media from Google Drive**

You can upload media, such as PDF, MP4, WAV, and JPG files from Google Drive, when you send [image](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/image-understanding), [video](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/video-understanding), [audio](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/audio-understanding), and [document](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/document-understanding) prompt requests.

## June 10, 2024

Feature **Experiment in the Vertex AI Studio login-free**

The Vertex AI Studio multi-model prompt designer can be accessed login-free. With this feature, prospective customers can use the Vertex AI Studio to test queries before deciding to sign up and create an account. To learn more about this experience, see [Vertex AI Studio console experiences](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/learn/overview#console-experiences) or to access the console directly go to [Vertex AI Studio](https://console.cloud.google.com/vertex-ai/generative/multimodal/create/text?_ga=2.160908414.2102276437.1718032367-729345960.1717709985).

## May 31, 2024

Change **Generative AI on Vertex AI Regional APIs**

Generative AI on Vertex AI regional APIs are available in the [following three regions](https://cloud.google.com/vertex-ai/generative-ai/docs/learn/locations):

- `us-east5`
- `me-central1`
- `me-central2`

<br />

Change **Anthropic Claude 3.0 Opus model**

The [Anthropic Claude 3.0 Opus](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/partner-models/use-claude#claude-opus) model is [Generally Available](https://cloud.google.com/products#product-launch-stages). To learn more, see its [model card in Model Garden](https://console.cloud.google.com/vertex-ai/publishers/anthropic/model-garden/claude-3-opus).

## May 28, 2024

Feature Gemini models support the `frequencyPenalty` and `presencePenalty` parameters. Use `frequencyPenalty` to control the probability of repeated text in a response. Use `presencePenalty` to control the probability of generating more diverse content. For more information, see [Gemini model parameters](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-reference/gemini#parameters).

## May 24, 2024

Announcement The Gemini 1.5 Pro (`gemini-1.5-pro-001`) and Gemini 1.5 Flash (`gemini-1.5-flash-001`) models are [Generally Available](https://cloud.google.com/products?e=13802955&hl=en#product-launch-stages). For more information, see [Google models](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/learn/models), [Overview of the Gemini API](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/overview), and [Send multimodal prompt requests](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/send-multimodal-prompts).

## May 20, 2024

Feature The following models have been added to [Model Garden](https://console.cloud.google.com/vertex-ai/model-garden):

- **E5**: A text embedding model series that can be served with a GPU or CPU.
- **Instant ID**: An identity preserving text-to-image generation model.
- **Stable Diffusion XL lightning**: A text-to-image generation model that is based on SDXL but requires fewer inference iterations.

To see a list of all available models, see [Explore models in Model Garden](https://docs.cloud.google.com/vertex-ai/docs/start/explore-models).

## May 14, 2024

Announcement **Gemini 1.5 Flash (Preview)**

[Gemini 1.5 Flash](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/send-multimodal-prompts) ([`gemini-1.5-flash-preview-0514`](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versioning#preview-models)) is available in [Preview](https://cloud.google.com/products?e=13802955&hl=en#product-launch-stages). Gemini 1.5 Flash is a multimodal model designed for fast, high volume, cost-effective text generation and chat applications. It can analyze text, code, audio, PDF, video, and video with audio.
Announcement **PaliGemma model**

The [PaliGemma](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/open-models/use-gemma) model is available. PaliGemma is a lightweight open model that's part of the Google Gemma model family. It's the Gemma model family's best model option for image captioning tasks and visual question and answering tasks. Gemma models are based on Gemini models and intended to be extended by customers.
Feature **Grounding Gemini with Google Search is GA**

The Gemini API Grounding with Google Search feature is available in [GA](https://cloud.google.com/products#product-launch-stages). This is available for Gemini 1.0 Pro models. To learn more about model grounding, see [Grounding with Google Search](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/ground-gemini#web-ground-gemini).
Feature **Batch prediction support for Gemini**

Batch prediction is available for Gemini in [preview](https://cloud.google.com/products#product-launch-stages). Available Gemini models include Gemini 1.0 Pro, Gemini 1.5 Pro, and Gemini 1.5 Flash. To get started with batch prediction, see [Get batch predictions for Gemini](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/batch-prediction-gemini).
Feature **New stable text embedding models**

The following text embedding models are available [GA](https://cloud.google.com/products#product-launch-stages):

- `text-embedding-004`
- `text-multilingual-embedding-002`

For details on how to use these models, see [Get text embeddings](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/embeddings/get-text-embeddings).

## April 18, 2024

Feature Meta's open weight [Llama 3 model](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/open-models/use-llama) is available in the Vertex AI Model Garden.

## April 11, 2024

Announcement **Anthropic Claude 3.0 Opus model**

The [Anthropic Claude 3.0 Opus](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/partner-models/use-claude#claude-opus) model is available in [Preview](https://cloud.google.com/products#product-launch-stages). The Claude 3.0 Opus model is an Anthropic partner model that you can use with Vertex AI. It's the most capable of the Anthropic models at performing complex tasks quickly. To learn more, see its [model card in Model Garden](https://console.cloud.google.com/vertex-ai/publishers/anthropic/model-garden/claude-3-opus).

## April 09, 2024

Change **Change in Imagen image generation version 006 (`imagegeneration@006`) `seed` field behavior**

For the new Imagen image generation model version 006 (`imagegeneration@006`) the `seed` field behavior has changed. For the v.006 model a digital watermark is enabled by default for image generation. To be able to use a `seed` value to get deterministic output you must disable digital watermark generation by setting the following `parameter`: `"addWatermark": false`.

For more information, see the [Imagen for image generation and editing API reference](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-reference/image-generation).
Feature **New Imagen on Vertex AI image editing model and features**

The 006 version of the Imagen 2 image editing model (`imagegeneration@006`) is now available. This model offers the following additional features:

- [Inpainting](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/edit-inpainting) - Add or remove content from a masked area of an image
- [Outpainting](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/edit-outpainting.md.txt) - Expand a masked area of an image
- [Product image editing](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/edit-product-image) - Identify and maintain a primary product while changing the background or product position

For more information, see [Model versions](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/model-versioning).
Feature **New Imagen on Vertex AI image generation model and features**

The 006 version of the Imagen 2 image generation model (`imagegeneration@006`) is now available. This model offers the following additional features:

- Additional aspect ratios (1:1, 3:4, 4:3, 9:16, 16:9)
- Digital watermark ([SynthID](https://deepmind.google/technologies/synthid/)) enabled by default
- Watermark verification\*
- New user-configurable safety features (safety setting, person/face setting)

For more information, see [Model versions](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/model-versioning) and [Generate images using text prompts](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/image/generate-images.md.txt).

\* The `seed` field can't be used while digital watermark is enabled.
Feature **Grounding Gemini and Grounding with Google Search**

The Gemini API now supports Grounding with Google Search in [Preview](https://cloud.google.com/products#product-launch-stages). Currently available for Gemini 1.0 Pro models.
Feature **Online Evaluation Service**

[Generative AI evaluation](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/evaluation-overview) supports [online evaluation](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/online-pipeline-services) in addition to [pipeline evaluation](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/side-by-side-eval). The list of supported evaluation metrics has also expanded. See [API reference](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/model-reference/evaluation) and [SDK reference](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/online-pipeline-services).
Feature **Supervised Tuning for Gemini**

[Supervised tuning](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/models/tune-gemini-overview) is available for the `gemini-1.0-pro-002 model`.
Change **Generative AI Knowledge Base**

The [Jump Start Solution: Generative AI Knowledge Base](https://console.cloud.google.com/products/solutions/details/generative-ai-knowledge-base) demonstrates how to build a simple chatbot with business- and domain-specific knowledge.
Feature **New text embedding models**

The following text embedding models are now in [Preview](https://cloud.google.com/products#product-launch-stages).

- `text-embedding-preview-0409`
- `text-multilingual-embedding-preview-0409`

When evaluated using the [MTEB benchmarks](https://huggingface.co/spaces/mteb/leaderboard), these models produce better embeddings compared to previous versions. The new models also offer [dynamic embedding sizes](https://arxiv.org/pdf/2205.13147.pdf), which you can use to output smaller embedding dimensions, with minor performance loss, to save on computing and storage costs.

For details on how to use these models, refer to the [public documentation](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/embeddings/get-text-embeddings#latest_models) and try out our [Colab](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/official/generative_ai/text_embedding_new_api.ipynb).
Feature **System instructions**

[System instructions](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/send-chat-prompts-gemini#system-instructions) are supported in [Preview](https://cloud.google.com/products#product-launch-stages) by the Gemini 1.0 Pro (stable version `gemini-1.0-pro-002` only) and Gemini 1.5 Pro ([Preview](https://cloud.google.com/products#product-launch-stages)) multimodal models. Use system instructions to guide model behavior based on your specific needs and use cases. For more information, see [System instructions examples](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/send-chat-prompts-gemini#system_instructions_examples).
Change **Regional APIs**

- Regional APIs are [available in 11 new countries](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/learn/locations#available-regions) for Gemini, Imagen, and embeddings.
- [US and EU](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/learn/locations) have machine-learning processing boundaries for the `gemini-1.0-pro-001`, `gemini-1.0-pro-002`, `gemini-1.0-pro-vision-001`, and `imagegeneration@005` models.
Feature **Text translation**

[Translate text in Vertex AI Studio](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/translate/translate-text) is available in [Preview](https://cloud.google.com/products#product-launch-stages).
Announcement **Gemini 1.0 Pro stable version 002**

The 002 version of the Gemini 1.0 Pro multimodal model (`gemini-1.0-pro-002`) is available. For more information about stable versions of Gemini models, see [Gemini model versions and lifecycle](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versioning#stable-version).
Feature **Generative AI on Vertex AI security control update**

[Security controls](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/security-controls) are available for the online prediction feature for Gemini 1.0 Pro and Gemini 1.0 Pro Vision.
Change **Vertex AI Studio features and updates**

- The Vertex AI Studio supports side-by-side comparison to allow users to compare up to 3 prompts in a side-by-side view.
- The Vertex AI Studio supports rapid evaluation in console and the ability to upload a ground truth response (or a model response to try to emulate).

To learn more, see [Try your prompts in Vertex AI Studio](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/start/quickstarts/quickstart)
Announcement **CodeGemma model**

The [CodeGemma](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/open-models/use-gemma) model is available. CodeGemma is a lightweight open model that's part of the Google Gemma model family. CodeGemma is the Gemma model family's code generation and code completion offering. Gemma models are based on Gemini models and intended to be extended by customers.
Announcement **Gemini 1.5 Pro (Preview)**

[Gemini 1.5 Pro](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/multimodal/send-multimodal-prompts) ([`gemini-1.5-pro-preview-0409`](https://docs.cloud.google.com/vertex-ai/generative-ai/docs/learn/model-versioning#preview-models)) is available in [Preview](https://cloud.google.com/products#product-launch-stages). Gemini 1.5 Pro is a multimodal model that analyzes text, code, audio, PDF, video, and video with audio.

## April 02, 2024

Feature [Model Garden](https://console.cloud.google.com/vertex-ai/model-garden) supports all [Text Generation Inference](https://huggingface.co/docs/text-generation-inference/en/index) supported models in [HuggingFace](https://huggingface.co/):

- Verified deployment settings for about 400 Hugging Face text generation models (including [google/gemma-7b-it](https://console.cloud.google.com/vertex-ai/publishers/google/model-garden/gemma-7b-it;action=deploy;hfSource=true), [meta-llama/Llama-2-7b-chat-hf](https://console.cloud.google.com/vertex-ai/publishers/meta-llama/model-garden/Llama-2-7b-chat-hf;action=deploy;hfSource=true), and [mistralai/Mistral-7B-v0.1](https://console.cloud.google.com/vertex-ai/publishers/mistralai/model-garden/Mistral-7B-v0.1;action=deploy;hfSource=true)).
- Other Hugging Face text generation models have unverified deployment settings that are auto generated.

## March 29, 2024

Change The MedLM-large model infrastructure has been upgraded to improve
latency and stability. Responses from the model might be slightly different.