Gemini Enterprise Agent Platform provides Docker container images that you run as *prebuilt
containers* for serving inferences and [explanations](https://docs.cloud.google.com/vertex-ai/docs/explainable-ai/overview.md.txt) from trained model
artifacts. These containers, which are organized by machine learning (ML)
framework and framework version, provide [HTTP inference
servers](https://docs.cloud.google.com/gemini-enterprise-agent-platform/machine-learning/predictions/custom-container-requirements#server) that you can use to
serve inferences with minimal configuration. In many cases, using a prebuilt
container is simpler than [creating your own custom container for
inference](https://docs.cloud.google.com/gemini-enterprise-agent-platform/machine-learning/predictions/use-custom-container).

This document lists the prebuilt containers for inferences and explanations,
and it describes how to use them with model artifacts that you [created using
Agent Platform's custom training
functionality](https://docs.cloud.google.com/gemini-enterprise-agent-platform/machine-learning/training/code-requirements) or model artifacts that you
created outside of Agent Platform.

## Support policy and schedule

Gemini Enterprise Agent Platform supports each framework version based on a schedule to
minimize security vulnerabilities. Review the
[Support policy schedule](https://docs.cloud.google.com/gemini-enterprise-agent-platform/machine-learning/framework-support-policy.md.txt#support_policy_schedule) to understand the implications of
the end-of-support and end-of-availability dates.

## Available container images

Each of the following container images is available in several
Artifact Registry repositories, which [store data in various
locations](https://docs.cloud.google.com/artifact-registry/docs/repo-locations). You can use any of
the URIs for an image when you perform custom training; each provides the same
container image. If you use the Google Cloud console to create a
[`Model`](https://docs.cloud.google.com/gemini-enterprise-agent-platform/reference/rest/v1/projects.locations.models.md.txt) resource,
the Google Cloud console selects the URI that best matches the [location where
you are using Agent Platform](https://docs.cloud.google.com/gemini-enterprise-agent-platform/machine-learning/general/locations) in order to reduce
latency.

> [!NOTE]
> **Note:** Using image names without the `latest` tag isn't supported. You must use an image with the `latest` tag.

### TensorFlow

#### Available TensorFlow container images (Click to expand)

| ML framework version | Supported accelerators (and CUDA version, if applicable) | End of patch and support date | End of availability | Supported images |
|---|---|---|---|---|
| 2.15 | CPU only | July 14, 2026 | Jan 14, 2027 | - `us-docker.pkg.dev/vertex-ai/prediction/tf2-cpu.2-15:latest` - `europe-docker.pkg.dev/vertex-ai/prediction/tf2-cpu.2-15:latest` - `asia-docker.pkg.dev/vertex-ai/prediction/tf2-cpu.2-15:latest` |
| 2.15 | GPU (CUDA 12.x) | July 14, 2026 | Jan 14, 2027 | - `us-docker.pkg.dev/vertex-ai/prediction/tf2-gpu.2-15:latest` - `europe-docker.pkg.dev/vertex-ai/prediction/tf2-gpu.2-15:latest` - `asia-docker.pkg.dev/vertex-ai/prediction/tf2-gpu.2-15:latest` |
| 2.14 | CPU only | Jan 14, 2026 | Jan 14, 2027 | - `us-docker.pkg.dev/vertex-ai/prediction/tf2-cpu.2-14:latest` - `europe-docker.pkg.dev/vertex-ai/prediction/tf2-cpu.2-14:latest` - `asia-docker.pkg.dev/vertex-ai/prediction/tf2-cpu.2-14:latest` |
| 2.14 | GPU (CUDA 12.x) | Jan 14, 2026 | Jan 14, 2027 | - `us-docker.pkg.dev/vertex-ai/prediction/tf2-gpu.2-14:latest` - `europe-docker.pkg.dev/vertex-ai/prediction/tf2-gpu.2-14:latest` - `asia-docker.pkg.dev/vertex-ai/prediction/tf2-gpu.2-14:latest` |
| 2.13 | CPU only | Nov 28, 2024 | Nov 28, 2025 |   |
| 2.13 | GPU (CUDA 12.x) | Nov 28, 2024 | Nov 28, 2025 |   |
| 2.12 | CPU only | June 30, 2024 | June 30, 2025 |   |
| 2.12 | GPU (CUDA 11.x) | June 30, 2024 | June 30, 2025 |   |
| 2.11 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.11 | GPU (CUDA 11.x) | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.10 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.10 | GPU (CUDA 11.x) | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.9 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.9 | GPU (CUDA 11.x) | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.8 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.8 | GPU (CUDA 11.x) | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.7 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.7 | GPU (CUDA 11.x) | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.6 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.6 | GPU (CUDA 11.x) | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.5 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.5 | GPU (CUDA 11.x) | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.4 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.4 | GPU (CUDA 11.x) | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.3 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.3 | GPU (CUDA 11.x) | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.2 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.2 | GPU (CUDA 11.x) | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.1 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.1 | GPU (CUDA 11.x) | Nov 15, 2023 | Nov 15, 2024 |   |
| 1.15 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 1.15 | GPU (CUDA 11.x) | Nov 15, 2023 | Nov 15, 2024 |   |

### Optimized TensorFlow runtime

The following container images use the optimized TensorFlow runtime. For
more information, see [Use the optimized TensorFlow runtime](https://docs.cloud.google.com/gemini-enterprise-agent-platform/machine-learning/predictions/optimized-tensorflow-runtime.md.txt).

#### Available optimized TensorFlow runtime container images (Click to expand)

| ML framework version | Supported accelerators (and CUDA version, if applicable) | End of patch and support date | End of availability | Supported images |
|---|---|---|---|---|
| nightly | CPU only | Not applicable | Not applicable | - `us-docker.pkg.dev/vertex-ai-restricted/prediction/tf_opt-cpu.nightly:latest` - `europe-docker.pkg.dev/vertex-ai-restricted/prediction/tf_opt-cpu.nightly:latest` - `asia-docker.pkg.dev/vertex-ai-restricted/prediction/tf_opt-cpu.nightly:latest` |
| nightly | GPU (CUDA 12.x) | Not applicable | Not applicable | - `us-docker.pkg.dev/vertex-ai-restricted/prediction/tf_opt-gpu.nightly:latest` - `europe-docker.pkg.dev/vertex-ai-restricted/prediction/tf_opt-gpu.nightly:latest` - `asia-docker.pkg.dev/vertex-ai-restricted/prediction/tf_opt-gpu.nightly:latest` |
| nightly | Cloud TPU | Not applicable | Not applicable | - `us-docker.pkg.dev/vertex-ai-restricted/prediction/tf_opt-tpu.nightly:latest` - `europe-docker.pkg.dev/vertex-ai-restricted/prediction/tf_opt-tpu.nightly:latest` - `asia-docker.pkg.dev/vertex-ai-restricted/prediction/tf_opt-tpu.nightly:latest` |
| 2.17 | CPU only | Jul 11, 2024 | Jul 11, 2025 | - `us-docker.pkg.dev/vertex-ai-restricted/prediction/tf_opt-cpu.2-17:latest` - `europe-docker.pkg.dev/vertex-ai-restricted/prediction/tf_opt-cpu.2-17:latest` - `asia-docker.pkg.dev/vertex-ai-restricted/prediction/tf_opt-cpu.2-17:latest` |
| 2.17 | GPU (CUDA 12.x) | Jul 11, 2024 | Jul 11, 2025 | - `us-docker.pkg.dev/vertex-ai-restricted/prediction/tf_opt-gpu.2-17:latest` - `europe-docker.pkg.dev/vertex-ai-restricted/prediction/tf_opt-gpu.2-17:latest` - `asia-docker.pkg.dev/vertex-ai-restricted/prediction/tf_opt-gpu.2-17:latest` |
| 2.17 | Cloud TPU | Jul 11, 2024 | Jul 11, 2025 | - `us-docker.pkg.dev/vertex-ai-restricted/prediction/tf_opt-tpu.2-17:latest` - `europe-docker.pkg.dev/vertex-ai-restricted/prediction/tf_opt-tpu.2-17:latest` - `asia-docker.pkg.dev/vertex-ai-restricted/prediction/tf_opt-tpu.2-17:latest` |
| 2.16 | CPU only | Apr 26, 2024 | Apr 26, 2025 |   |
| 2.16 | GPU (CUDA 12.x) | Apr 26, 2024 | Apr 26, 2025 |   |
| 2.16 | Cloud TPU | Apr 26, 2024 | Apr 26, 2025 |   |
| 2.15 | CPU only | Aug 15, 2024 | Aug 15, 2025 |   |
| 2.15 | GPU (CUDA 12.x) | Aug 15, 2024 | Aug 15, 2025 |   |
| 2.15 | Cloud TPU | Aug 15, 2024 | Aug 15, 2025 |   |
| 2.14 | CPU only | Aug 15, 2024 | Aug 15, 2025 |   |
| 2.14 | GPU (CUDA 12.x) | Aug 15, 2024 | Aug 15, 2025 |   |
| 2.13 | CPU only | Aug 15, 2024 | Aug 15, 2025 |   |
| 2.13 | GPU (CUDA 11.x) | Aug 15, 2024 | Aug 15, 2025 |   |
| 2.12 | CPU only | May 15, 2024 | May 15, 2025 |   |
| 2.12 | GPU (CUDA 11.x) | May 15, 2024 | May 15, 2025 |   |
| 2.11 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.11 | GPU (CUDA 11.x) | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.10 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.10 | GPU (CUDA 11.x) | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.9 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.9 | GPU (CUDA 11.x) | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.8 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 2.8 | GPU (CUDA 11.x) | Nov 15, 2023 | Nov 15, 2024 |   |

<br />

### PyTorch

#### Available PyTorch container images (Click to expand)

| ML framework version | Supported accelerators (and CUDA version, if applicable) | End of patch and support date | End of availability | Supported images |
|---|---|---|---|---|
| 2.4 (Python 3.9) | CPU only | July 14, 2026 | Jan 14, 2027 | - `us-docker.pkg.dev/vertex-ai/prediction/pytorch-cpu.2-4:latest` - `europe-docker.pkg.dev/vertex-ai/prediction/pytorch-cpu.2-4:latest` - `asia-docker.pkg.dev/vertex-ai/prediction/pytorch-cpu.2-4:latest` |
| 2.4 (Python 3.9) | GPU (CUDA 12.x) | July 14, 2026 | Jan 14, 2027 | - `us-docker.pkg.dev/vertex-ai/prediction/pytorch-gpu.2-4:latest` - `europe-docker.pkg.dev/vertex-ai/prediction/pytorch-gpu.2-4:latest` - `asia-docker.pkg.dev/vertex-ai/prediction/pytorch-gpu.2-4:latest` |
| 2.4 (Python 3.9) | Cloud TPU | July 14, 2026 | Jan 14, 2027 | - `us-docker.pkg.dev/vertex-ai/prediction/pytorch-tpu.2-4:latest` - `europe-docker.pkg.dev/vertex-ai/prediction/pytorch-tpu.2-4:latest` - `asia-docker.pkg.dev/vertex-ai/prediction/pytorch-tpu.2-4:latest` |
| 2.3 (Python 3.9) | CPU only | Jan 14, 2026 | Jan 14, 2027 | - `us-docker.pkg.dev/vertex-ai/prediction/pytorch-cpu.2-3:latest` - `europe-docker.pkg.dev/vertex-ai/prediction/pytorch-cpu.2-3:latest` - `asia-docker.pkg.dev/vertex-ai/prediction/pytorch-cpu.2-3:latest` |
| 2.3 (Python 3.9) | GPU (CUDA 12.x) | Jan 14, 2026 | Jan 14, 2027 | - `us-docker.pkg.dev/vertex-ai/prediction/pytorch-gpu.2-3:latest` - `europe-docker.pkg.dev/vertex-ai/prediction/pytorch-gpu.2-3:latest` - `asia-docker.pkg.dev/vertex-ai/prediction/pytorch-gpu.2-3:latest` |
| 2.3 (Python 3.9) | Cloud TPU | Jan 14, 2026 | Jan 14, 2027 | - `us-docker.pkg.dev/vertex-ai/prediction/pytorch-tpu.2-3:latest` - `europe-docker.pkg.dev/vertex-ai/prediction/pytorch-tpu.2-3:latest` - `asia-docker.pkg.dev/vertex-ai/prediction/pytorch-tpu.2-3:latest` |
| 2.2 (Python 3.9) | CPU only | Jan 14, 2026 | Jan 14, 2027 | - `us-docker.pkg.dev/vertex-ai/prediction/pytorch-cpu.2-2:latest` - `europe-docker.pkg.dev/vertex-ai/prediction/pytorch-cpu.2-2:latest` - `asia-docker.pkg.dev/vertex-ai/prediction/pytorch-cpu.2-2:latest` |
| 2.2 (Python 3.9) | GPU (CUDA 12.x) | Jan 14, 2026 | Jan 14, 2027 | - `us-docker.pkg.dev/vertex-ai/prediction/pytorch-gpu.2-2:latest` - `europe-docker.pkg.dev/vertex-ai/prediction/pytorch-gpu.2-2:latest` - `asia-docker.pkg.dev/vertex-ai/prediction/pytorch-gpu.2-2:latest` |
| 2.2 (Python 3.9) | Cloud TPU | Jan 14, 2026 | Jan 14, 2027 | - `us-docker.pkg.dev/vertex-ai/prediction/pytorch-tpu.2-2:latest` - `europe-docker.pkg.dev/vertex-ai/prediction/pytorch-tpu.2-2:latest` - `asia-docker.pkg.dev/vertex-ai/prediction/pytorch-tpu.2-2:latest` |
| 2.1 (Python 3.9) | CPU only | Dec 01, 2024 | Dec 01, 2025 |   |
| 2.1 (Python 3.9) | GPU (CUDA 12.x) | Dec 01, 2024 | Dec 01, 2025 |   |
| 2.1 (Python 3.9) | Cloud TPU | Dec 01, 2024 | Dec 01, 2025 |   |
| 2.0 (Python 3.9) | CPU only | Jul 27, 2024 | Jul 27, 2025 |   |
| 2.0 (Python 3.9) | GPU (CUDA 11.x) | Jul 27, 2024 | Jul 27, 2025 |   |
| 1.13 (Python 3.8) | CPU only | May 15, 2024 | May 15, 2025 |   |
| 1.13 (Python 3.8) | GPU (CUDA 11.x) | May 15, 2024 | May 15, 2025 |   |
| 1.12 | CPU only | May 15, 2024 | May 15, 2025 |   |
| 1.12 | GPU (CUDA 11.x) | May 15, 2024 | May 15, 2025 |   |
| 1.11 | CPU only | May 15, 2024 | May 15, 2025 |   |
| 1.11 | GPU (CUDA 11.x) | May 15, 2024 | May 15, 2025 |   |

### scikit-learn

#### Available scikit-learn container images (Click to expand)

| ML framework version | Supported accelerators (and CUDA version, if applicable) | End of patch and support date | End of availability | Supported images |
|---|---|---|---|---|
| 1.6 (Python 3.10) | CPU only | Oct 14, 2026 | Oct 14, 2027 | - `us-docker.pkg.dev/vertex-ai/prediction/sklearn-cpu.1-6:latest` - `europe-docker.pkg.dev/vertex-ai/prediction/sklearn-cpu.1-6:latest` - `asia-docker.pkg.dev/vertex-ai/prediction/sklearn-cpu.1-6:latest` |
| 1.5 (Python 3.10) | CPU only | July 14, 2026 | Jan 14, 2027 | - `us-docker.pkg.dev/vertex-ai/prediction/sklearn-cpu.1-5:latest` - `europe-docker.pkg.dev/vertex-ai/prediction/sklearn-cpu.1-5:latest` - `asia-docker.pkg.dev/vertex-ai/prediction/sklearn-cpu.1-5:latest` |
| 1.4 (Python 3.10) | CPU only | Jan 14, 2026 | Jan 14, 2027 | - `us-docker.pkg.dev/vertex-ai/prediction/sklearn-cpu.1-4:latest` - `europe-docker.pkg.dev/vertex-ai/prediction/sklearn-cpu.1-4:latest` - `asia-docker.pkg.dev/vertex-ai/prediction/sklearn-cpu.1-4:latest` |
| 1.3 (Python 3.10) | CPU only | Nov 28, 2024 | Nov 28, 2025 |   |
| 1.2 (Python 3.10) | CPU only | June 30, 2024 | June 30, 2025 |   |
| 1.0 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 0.24 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 0.23 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 0.22 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 0.20 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |

### XGBoost

#### Available XGBoost container images (Click to expand)

| ML framework version | Supported accelerators (and CUDA version, if applicable) | End of patch and support date | End of availability | Supported images |
|---|---|---|---|---|
| 2.1 (Python 3.10) | CPU only | July 14, 2026 | Jan 14, 2027 | - `us-docker.pkg.dev/vertex-ai/prediction/xgboost-cpu.2-1:latest` - `europe-docker.pkg.dev/vertex-ai/prediction/xgboost-cpu.2-1:latest` - `asia-docker.pkg.dev/vertex-ai/prediction/xgboost-cpu.2-1:latest` |
| 2.0 (Python 3.10) | CPU only | Jan 14, 2026 | Jan 14, 2027 | - `us-docker.pkg.dev/vertex-ai/prediction/xgboost-cpu.2-0:latest` - `europe-docker.pkg.dev/vertex-ai/prediction/xgboost-cpu.2-0:latest` - `asia-docker.pkg.dev/vertex-ai/prediction/xgboost-cpu.2-0:latest` |
| 1.7 (Python 3.10) | CPU only | June 30, 2024 | Dec 30, 2025 |   |
| 1.6 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 1.5 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 1.4 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 1.3 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 1.2 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 1.1 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 0.90 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |
| 0.82 | CPU only | Nov 15, 2023 | Nov 15, 2024 |   |

## Use a prebuilt container

You can specify a prebuilt container for inference when you
[create a custom `TrainingPipeline` resource that uploads a `Model`](https://docs.cloud.google.com/gemini-enterprise-agent-platform/machine-learning/training/create-training-pipeline.md.txt#custom-job-model-upload) or when
you [import model artifacts as a `Model`](https://docs.cloud.google.com/gemini-enterprise-agent-platform/machine-learning/model-registry/import-model.md.txt).

To use one of these prebuilt containers, you must save your model as one or
more *model artifacts* that comply with the requirements of the prebuilt
container. For more information, see
[Export model artifacts for inference](https://docs.cloud.google.com/gemini-enterprise-agent-platform/machine-learning/training/exporting-model-artifacts.md.txt).

The following notebooks demonstrate how to use a prebuilt container to serve
inferences.

| What do you want to do? | Notebook |
|---|---|
| Train and serve a TensorFlow model using a prebuilt container | [Custom training and online inference](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/official/custom/sdk-custom-image-classification-online.ipynb) |
| Serve a PyTorch model using a prebuilt container | [Serving PyTorch image models with prebuilt containers on Agent Platform](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/official/prediction/pytorch_image_classification_with_prebuilt_serving_containers.ipynb) |
| Serve a Stable Diffusion model using a prebuilt container | [Deploy and host a Stable Diffusion model on Agent Platform](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/community/vertex_endpoints/torchserve/dreambooth_stablediffusion.ipynb) |

## Notebooks

> [!NOTE]
> To learn more,
> run the "Serving PyTorch image models with prebuilt containers on " notebook in one of the following
> environments:
>
> [![](https://docs.cloud.google.com/static/vertex-ai/images/colab-logo-32px.png)Open in Colab](https://colab.research.google.com/github/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/official/prediction/pytorch_image_classification_with_prebuilt_serving_containers.ipynb)
>
>
> \|
>
> [![](https://docs.cloud.google.com/static/vertex-ai/images/colab-enterprise-logo-32px.png)Open in Colab Enterprise](https://console.cloud.google.com/agent-platform/colab/import/https%3A%2F%2Fraw.githubusercontent.com%2FGoogleCloudPlatform%2Fvertex-ai-samples%2Fmain%2Fnotebooks%2Fofficial%2Fprediction%2Fpytorch_image_classification_with_prebuilt_serving_containers.ipynb)
>
>
> \|
>
> [![](https://docs.cloud.google.com/static/vertex-ai/images/vertex-ai-workbench-logo-32px.png)Open
> in Agent Platform Workbench](https://console.cloud.google.com/agent-platform/workbench/deploy-notebook?download_url=https://raw.githubusercontent.com/GoogleCloudPlatform/vertex-ai-samples/main/notebooks/official/prediction/pytorch_image_classification_with_prebuilt_serving_containers.ipynb)
>
>
> \|
>
> [![](https://docs.cloud.google.com/static/vertex-ai/images/github-logo-32px.png)View on GitHub](https://github.com/GoogleCloudPlatform/vertex-ai-samples/blob/main/notebooks/official/prediction/pytorch_image_classification_with_prebuilt_serving_containers.ipynb)

## What's next

- Learn how to [deploy a model to an endpoint to serve
  inferences](https://docs.cloud.google.com/gemini-enterprise-agent-platform/machine-learning/predictions/deploy-model-api).