> ## Documentation Index
> Fetch the complete documentation index at: https://dripart-docs-partner-nodes-consolidate.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# BatchLatentsNode - ComfyUI Built-in Node Documentation

> The Batch Latents node combines multiple latent inputs into a single batch.

The Batch Latents node combines multiple latent inputs into a single batch. It takes a variable number of latent samples and merges them along the batch dimension so they can be processed together by subsequent nodes. The node also merges the batch index metadata of all inputs into the combined output.

Batch Latents（批处理潜空间）节点将多个潜空间输入合并为单个批次。它接收可变数量的潜空间样本，并沿批次维度将它们合并，以便后续节点可以一起处理。该节点还会将所有输入的批次索引元数据合并到组合输出中。

## Inputs

## 输入

| Parameter | Description                                                                                                                                                                                         | Data Type | Required | Range         |
| --------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | --------- | -------- | ------------- |
| `latent`  | A set of latent samples to be combined into a single batch. You must provide at least one latent, and you can add up to 50. The node automatically creates input slots as you connect more latents. | LATENT    | 是        | 1 到 50 inputs |

| 参数       | 描述                                                                  | 数据类型   | 是否必需 | 范围            |
| -------- | ------------------------------------------------------------------- | ------ | ---- | ------------- |
| `latent` | 要合并为单个批次的潜空间样本集合。您必须至少提供一个潜空间，并且最多可以添加 50 个。当您连接更多潜空间时，节点会自动创建输入插槽。 | LATENT | 是    | 1 到 50 inputs |

**Note:** You must provide at least one latent input for the node to function. The node automatically creates input slots as you connect more latents, up to a maximum of 50. Each input latent is reshaped to match the sample shape of the first latent before being combined, and any latent without batch index metadata is assigned a sequential batch index.

**注意：** 您必须至少提供一个潜空间输入，节点才能正常工作。当您连接更多潜空间时，节点会自动创建输入插槽，最多可达 50 个。每个输入潜空间在合并前都会调整形状，以匹配第一个潜空间的样本形状；任何没有批次索引元数据的潜空间都会被分配一个顺序批次索引。

## Outputs

## 输出

| Output Name | Description                                                                                                                       | Data Type |
| ----------- | --------------------------------------------------------------------------------------------------------------------------------- | --------- |
| `output`    | A single latent output containing all the input latents combined into one batch, together with their merged batch index metadata. | LATENT    |

| 输出名称     | 描述                                   | 数据类型   |
| -------- | ------------------------------------ | ------ |
| `output` | 包含所有输入潜空间合并为一个批次及合并后批次索引元数据的单一潜空间输出。 | LATENT |

> 本文档由 AI 生成。如果您发现任何错误或有改进建议，欢迎贡献！ [在 GitHub 上编辑](https://github.com/Comfy-Org/embedded-docs/blob/main/comfyui_embedded_docs/docs/BatchLatentsNode/zh.md)

***

**Source fingerprint (SHA-256):** `38df5e6cfa391e054c663af1cc55728d115cebfbb804e1c2c51dfc2aab37df47`