> For the complete documentation index, see [llms.txt](https://docs.cherryai.com.cn/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.cherryai.com.cn/docs/en-us/advanced-basic/knowledge-content/translation-workflow.md).

# Files, images, and long-form translation

\[Translation] It can handle directly entered text, or you can upload images or documents. Images will have their text recognized first, while documents are processed according to format and content; history can be favorited, reopened, and reused.

<figure><img src="https://1658303467-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F0Ut5BptC3t8CtSU1UWpM%2Fuploads%2FuAux5cNPlWeEphenKT5f%2Fclipboard.png?alt=media&amp;token=4875b86d-943e-4aea-85db-9085048edafe" alt="资料经过检索和 Agent 处理后形成文字、图片或译文的内容工作流图"><figcaption><p>Translation is part of the content workflow: first confirm the source materials, then unify terminology and check the final deliverable.</p></figcaption></figure>

### Choose input method

| Materials                  | Recommended practice                                   | Key checks                                                  |
| -------------------------- | ------------------------------------------------------ | ----------------------------------------------------------- |
| Short text                 | Paste directly                                         | Tone, proper nouns, context                                 |
| Screenshot or scanned copy | Upload an image and use OCR                            | Misrecognized characters, tables, and numbers               |
| PDF                        | Upload the file and choose whether to install BabelDOC | Two-column layout, fonts, tables, footnotes, and page count |
| Other long documents       | Upload file                                            | Heading hierarchy, paragraphs, footnotes, and formatting    |
| Markdown                   | Use preview to check                                   | Whether links, code blocks, and markup are preserved        |

#### Example: short text translation

Choose \[Auto-detect] as the source language and \[English] as the target language, enter Chinese text, then click \[Translate]. After the result appears, first verify the key meanings of terms such as "gray release," "monitoring metrics," and "rollback process," then copy it or save it to notes.

<figure><img src="https://3562065924-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F0Ut5BptC3t8CtSU1UWpM%2Fuploads%2Fgit-blob-ee99f863bdfef7403502cdc2b4ad505c0cf66069%2Fcherry-v2-guide-translation-result.png?alt=media" alt="中文发布方案翻译为英文的完整结果"><figcaption><p>① The original text is kept on the left; ② The translation is shown on the right. Before delivery, check gray release, monitoring metrics, and rollback process item by item.</p></figcaption></figure>

#### What to check after the result comes out

| Check item             | Example                                                             |
| ---------------------- | ------------------------------------------------------------------- |
| Key meaning            | "Gray release" cannot be translated literally based only on color   |
| Proper nouns           | Keep product names, team names, and feature names consistent        |
| Numbers and conditions | Dates, ratios, amounts, and negative words must not be omitted      |
| Tone                   | Notices, contracts, and marketing copy should use appropriate tones |

Short text can be checked sentence by sentence directly; for files and long texts, first spot-check headings, tables, numbers, and repeated terms, then decide whether to deliver the whole piece.

### Translate PDF while preserving layout

After uploading a PDF with extractable text, the page will detect BabelDOC. For first-time use, choose \[Install BabelDOC]; once the dependencies are ready, the left side previews the original text and the right side shows the translated PDF, with progress displayed for resource checking, parsing, translation, layout, and rendering.

{% stepper %}
{% step %}

#### 1. Upload the PDF and check dependencies

Click \[Drag in or click to upload image/document] to select a PDF. If the page indicates that BabelDOC is missing or outdated, install or update it; installation is centrally managed by Cherry Studio's \[Environment Dependencies].
{% endstep %}

{% step %}

#### 2. Select the translation model and target language

PDF translation requires the current model to be callable through the local API gateway. If the model is unavailable, go back to \[Settings] → \[Model Services] to check the connection; do not repeatedly install BabelDOC.
{% endstep %}

{% step %}

#### 3. Compare the two-column result

After translation is complete, compare the original text and translation page by page, focusing on tables, captions, headers and footers, footnotes, and line breaks. Click the download button to save the translated PDF separately.
{% endstep %}

{% step %}

#### 4. Continue from history

PDF translations will appear in \[Translation History] and \[Files]. In history, you can reopen the two-column preview, locate the translated file, or save the result again; when a record is deleted, the corresponding translated file will also be recycled.
{% endstep %}
{% endstepper %}

{% hint style="warning" %}
Scanned or image-only PDFs do not currently support layout-preserving translation. If the page indicates that OCR is needed, first recognize the image text into editable text, then use the normal translation workflow.
{% endhint %}

### Translation steps

{% stepper %}
{% step %}

#### 1. First confirm the source and target languages

Auto-detection is suitable for mixed or unknown languages; formal materials such as contracts and technical documents are recommended to be manually confirmed to reduce misjudgments.
{% endstep %}

{% step %}

#### 2. Choose a translation method

Use the configured model when you need consistency in context, tone, and terminology; ordinary short sentences can use a lighter method. Available options depend on the current page.
{% endstep %}

{% step %}

#### 3. Check by segments

First check names, numbers, dates, negative words, and units. For image translation, also verify the OCR result; if recognition is wrong, the following translation will also be wrong.
{% endstep %}

{% step %}

#### 4. Favorite, export, or reuse the result

In \[Translation History], view, favorite, and reuse text translations; PDF results can be previewed in two columns, opened, or saved separately. Before formal delivery, save the glossary and manual revisions in the project materials.
{% endstep %}
{% endstepper %}

#### Application case: preparing multilingual product materials

First organize the product name, feature names, and terms that must not be rewritten into a specification. Upload the Chinese manual, translate one chapter first and confirm the terminology, then process the full text; upload promotional images separately for OCR translation. After completion, put the confirmed terms and translations into the project workspace and let the Agent check titles, links, and numbers consistently, but do not let it rewrite legal statements on its own.

{% hint style="danger" %}
Medical, legal, financial, and safety-related content requires professional review. Automatic translation can improve efficiency, but it cannot replace review by an accountable human reviewer.
{% endhint %}


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.cherryai.com.cn/docs/en-us/advanced-basic/knowledge-content/translation-workflow.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.