> ## Documentation Index
> Fetch the complete documentation index at: https://docs.scitix.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Models Overview

> Learn about preset models, custom models, model versions, and model reference methods in Siverse.

Models is used to register reusable models on the platform, including Preset models maintained by the platform and Custom models that you register. After registration, models can be referenced in Develop, Training, or Inference for model loading, fine-tuning, evaluation, or online deployment.

Model management primarily records the location of model files, version, type, tags, description, and metadata, helping teams consistently find and reuse model assets.

## Applicable Scenarios

* Find platform-maintained public models and reference them in training or inference scenarios.
* Register training artifacts, proprietary models, or third-party models as models that can be referenced on the platform.
* Prepare model paths, versions, and metadata for inference services.

## Model Source

| Source | Description | Suitable Scenarios |
| - | - | - |
| Preset Models | Public models maintained by the platform that can be viewed, referenced, or deployed directly. | Use platform-preconfigured models such as Qwen and DeepSeek. |
| Custom Models | Register your own model files on the platform and specify the Version, storage location, and metadata. | Reuse training artifacts, self-developed models, or external models. |

## Model Information

The model list and details typically include the following information:

| Information | Description |
| - | - |
| Region and Cluster | The scope of resources where the model resides. When creating a development environment, training task, or inference service, select a region and cluster that can access the model. |
| Name and Version | Used to distinguish model assets and model versions. |
| Type | The model task type, such as text generation. |
| Storage Location | The storage type, storage resource, and path where the model files are located. |
| Tags and Description | Used to search for, categorize, and describe the intended use of the model. |
| Metadata | Additional information such as model architecture, data type, parameters, and metrics. |

## Usage Procedure

1. Select **Preset** or **Custom** based on the model source.
2. Find an existing model or register a new custom model.
3. Confirm the model name, version, storage path, and metadata.
4. Reference the model in development environments, training tasks, or inference services.

Models only registers the model asset itself. For inference service resource specifications, service access, scaling, and service operations and maintenance, see [Create an LLM inference service](/siverse/inference/create-llm-inference-services.md), [Manage LLM inference services](/siverse/inference/manage-llm-inference-services.md), and [Advanced inference service operations](/siverse/inference/advanced-inference-service-operations.md).

## Next Steps

* [Use Preset Models](/siverse/models/use-built-in-models.md)
* [Register and Manage Custom Models](/siverse/models/register-and-manage-custom-models.md)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.