AI-Powered Automatic Translation for Google Slides - n8n Workflow

Business Consultants and Sales Professionals needing to localize presentations rapidly.
Global Teams requiring accurate, high-volume presentation translation.
n8n automation developers looking for advanced examples of AI agents and recursive n8n workflow structures.
Users seeking robust n8n templates for Google Workspace automation.

Translating large presentations manually is time-consuming, prone to error, and often loses valuable context or disrupts formatting. This advanced n8n workflow solves this critical localization challenge by combining Google Workspace APIs (Slides, Drive) with a powerful AI Agent utilizing Google Gemini.

This n8n automation handles the complexity of text extraction, chunking, recursive processing, contextual translation (excluding proper nouns, links, etc.), and accurate text replacement within the slides. A key technical feature of this n8n workflow is the use of self-execution (the Execute Workflow n8n node) to manage the processing of potentially hundreds of text blocks efficiently, ensuring the final translated presentation retains its original structure and design. This is one of the most powerful n8n templates available for high-fidelity multilingual content localization.

This n8n workflow operates in two main segments: initialization/chunking, and recursive translation/replacement.

1. Setup and Trigger: The primary n8n workflow is initiated manually. The user first configures the target language (e.g., zh-CN for Simplified Chinese) in the Set Language n8n node, using the ISO-639 code.


2. Duplication: The Google Drive n8n node duplicates the original Google Slides presentation, securing the source material.


3. Extraction: The Google Slides n8n node retrieves all raw content data. A custom Extract Text Code n8n node meticulously parses the complex slide objects, isolating every text block along with the corresponding slide identifier (objectId).


4. Recursion and Chunking: The flow utilizes the Split Out n8n node and the Execute Workflow n8n node to recursively call the same n8n workflow. This passes essential variables (lang, objectId, presentationId) for parallel or sequential processing of individual translation tasks, enabling effective throttling.


5. AI Agent Execution (Recursive Segment): The recursive segment is triggered by the When Executed by Another Workflow n8n trigger. It feeds the text chunk and target language to the Translation expert AI Agent.


6. Contextual Translation: The Translation expert AI Agent, powered by the Google Gemini Chat Model, uses a detailed system prompt defining specific translation boundaries (e.g., ignoring URLs, phone numbers, and company names) and utilizes the Translate language (Google Translate) n8n tool to perform the translation.


7. Text Replacement: The AI Agent instructs the Translate Google Slides n8n node to use the original text as a search key and the AI-generated translation as the replacement value. It uses the received objectId to accurately perform the replacement on the correct slide within the duplicated presentation.


8. Flow Control: Wait n8n nodes (set to 3 and 10 seconds) are strategically placed to manage Google API rate limits and ensure the stability of this complex n8n workflow.

To deploy this comprehensive n8n workflow, follow these setup steps:

1. Import: Import the provided n8n workflow JSON into your n8n instance.


2. Google Workspace Credentials: This n8n workflow requires credentials for Google Drive, Google Slides, and Google Translate (for the Translate language n8n node). Ensure all OAuth2 API credentials are correctly configured for the respective n8n nodes.


3. AI Credentials: Configure your Google Gemini (PaLM) API credentials for the Google Gemini Chat Model n8n node.


4. Target Language Setup: In the Set Language n8n node, update the lang value (currently zh-CN) to the ISO-639 code of your desired target language.


5. Source Presentation: In the Duplicate presentation n8n node, replace the placeholder file ID (XXXXXXX) with the actual ID of the Google Slides presentation you intend to translate.


6. Recursive Link Check: After saving the n8n workflow, verify that the Execute Workflow n8n node is correctly pointing to the unique ID of this newly saved n8n workflow itself.

When clicking ‘Execute workflow’ (n8n trigger): The manual trigger that starts the entire n8n automation process.
Set Language (Set n8n node): Configures the critical parameter—the target language code (ISO-639) for the translation.
Duplicate presentation (Google Drive n8n node): Performs a copy operation on the source presentation, ensuring non-destructive processing.
Extract Text (Code n8n node): Contains custom JavaScript essential for traversing the complex Google Slides API response structure and extracting plain text blocks along with their corresponding slide objectId. This data is crucial for the subsequent replacement steps in the n8n workflow.
Execute Workflow (n8n node): Facilitates recursion by executing a copy of this same n8n workflow, passing necessary slide and translation parameters for chunked processing.
When Executed by Another Workflow (n8n trigger): The entry point for the recursive segment of the n8n workflow, receiving the individual slide data for translation.
Translation expert (AI Agent n8n node): The core intelligence. This n8n node uses a detailed system prompt to guide the Gemini LLM to utilize the configured translation tools while adhering to strict rules regarding which content to translate and which to preserve.
Translate Google Slides (Google Slides Tool n8n node): An AI Tool utilized by the agent to perform the find-and-replace operation on the specific slide identified by its objectId and presentationId within the n8n workflow.