GPT-4o Multimodal AI Chat Assistant (Text, Images, PDFs) n8n

Chat with text, images, and PDFs using GPT-4o—built inside n8n

This GPT-4o Multimodal AI Chat Assistant n8n workflow lets you create a smart AI chat that can understand text, images, and PDFs. Upload files in the chat, ask questions, and get contextual answers powered by OpenAI’s GPT-4o multimodal model.

What this workflow does

• Starts a session with the n8n hosted chat UI via a chat trigger node.
• Accepts user input as either text or an uploaded file.
• Checks whether a file was included using conditional logic (e.g., an if step).
• If an image is uploaded, converts it to base64 and analyzes it using GPT-4o vision capabilities.
• Generates a natural-language description of the image (or handles mixed media) and responds to the user’s question in context.
• Keeps conversation state with a memory buffer so follow-up questions work intelligently.
• Uses GPT-4o to handle text-only and mixed media input seamlessly.

Use cases

• PDF explainer bot for documents you want users to ask questions about.
• Internal knowledge chat that supports media-assisted explanations (e.g., screenshots, diagrams, document uploads).
• AI-driven support assistant embedded in a website, where customers can upload images or PDFs to speed up troubleshooting.
• Personal assistant style chat with intelligent follow-ups across a conversation thread.

Technical details

• n8n-nodes-langchainchat chat trigger (hosted chat UI)
• n8nn8n-nodes-langchainagent / n8nn8n-nodes-langchainchain llm to route prompts to the LLM
• n8nn8n-nodes-langchainopen ai for OpenAI GPT-4o access (requires your API key)
• if + supporting nodes (including sticky note) to manage file detection and flow
• Requirement: configure File Upload on the chat