Smart PDF Summarizer with AI and Audio Playback - n8n Workflow

Content professionals needing fast PDF analysis.
Developers integrating AI summarization services via a webhook n8n trigger.
Users looking for advanced examples of n8n templates involving LangChain and custom HTTP polling.
Teams requiring automated text-to-speech functionality within their n8n node sequences.

Managing large volumes of PDF data can be time-consuming. This intelligent n8n workflow solves the problem by providing a fast, scalable solution for processing documents. It leverages the speed of Groq's language models for instant summarization and then converts that summary into an accessible audio file using a dedicated Text-to-Speech (TTS) service.

This n8n template demonstrates advanced capabilities, including binary file handling (PDF extraction), complex AI orchestration using the LangChain n8n node, custom code execution for data cleanup, and external API polling (using a standard n8n node for HTTP requests) to handle asynchronous operations. If you are looking for a comprehensive example of an AI-driven n8n workflow, this is an essential resource.

The entire process is initiated by an external request to the designated n8n trigger:

1. Webhook Trigger: The n8n workflow starts when a PDF file is POSTed to the Webhook n8n trigger endpoint (path: /summarize-pdf).


2. PDF Text Extraction: The Extract from File n8n node takes the binary file and extracts the raw text content from the PDF.


3. Data Preparation: A Python Code n8n node cleans up the text and prepares it for the AI Agent.


4. AI Summarization: The AI Agent n8n node, configured with the high-speed Groq Chat Model (openai/gpt-oss-20b) and Simple Memory, takes the full text and executes the instruction to summarize it into 4-5 lines.


5. TTS Initiation: The resulting summary text is captured by a Set n8n node and then sent to the Qwen TTS service using an HTTP Request n8n node to request audio generation. This request returns an event ID.


6. TTS Polling: A subsequent HTTP Request n8n node (TTS Poll) uses the event ID to continuously check the status of the audio generation until the final audio URL is available.


7. URL Parsing: A final JavaScript Code n8n node (Extract Audio URL) is crucial here, as it parses the complex server-sent events (SSE) response from the polling node to reliably locate and extract the specific audioUrl.


8. Response: The Edit Fields n8n node merges the summary and the new audio URL, which is then sent back immediately to the requester via the Respond to Webhook n8n node.

To set up this powerful n8n workflow, follow these steps:

1. Import the n8n template: Copy the provided JSON data and import it into your n8n instance via the 'Workflows' section by clicking 'New' and then 'Import from JSON'.


2. Setup Credentials: You need to configure credentials for the Groq service. Click on the Groq Chat Model1 n8n node and provide your Groq API key in the credentials setup.


3. Activate the Webhook: The Webhook n8n trigger is the starting point. Save the n8n workflow and click 'Activate'. Copy the unique testing or production webhook URL.


4. Testing: Test the n8n workflow by sending a POST request containing a PDF file to the copied webhook URL. Ensure your request handles binary file upload (e.g., using form-data).

Webhook (n8n trigger): Sets up the listening endpoint (/summarize-pdf) to receive the PDF file via a POST request, initiating the n8n workflow.
Extract from File (n8n node): Essential for pdf_processing, this node handles the binary data from the webhook and converts the PDF content into extractable text data.
AI Agent1 (n8n node): The core intelligence. It uses the Groq model (openai/gpt-oss-20b) to execute the summarization prompt. It is configured with Simple Memory and the Groq Chat Model1 n8n node as its Language Model.
TTS Request/TTS Poll (HTTP Request n8n node): Manages the interaction with the external Qwen TTS service. The first node sends the summary, and the second polls the service using the generated event ID to retrieve the completed audio file.
Extract Audio URL (Code n8n node): A critical n8n node written in JavaScript. Its function is to robustly parse the streaming or raw JSON output from the TTS polling service to extract the final usable audioUrl needed for the final response.
Respond with Both (n8n node): The final action that sends a structured JSON response back to the client, containing both the AI-generated summary and the generated audioUrl from the n8n workflow.