Telegram AI Chatbot with LangChain, GPT-4 & DALL-E

This n8n workflow creates a sophisticated Telegram AI chatbot powered by LangChain nodes. It starts by listening for incoming Telegram messages, then routes them to a Conversation Agent configured with a custom system prompt for formatted replies. The agent leverages GPT-4 (1106-preview model) for generating context-aware responses, maintains user-specific conversation history via Window Buffer Memory (10-message window keyed by Telegram chat ID), and integrates a custom DALL-E 3 tool for on-demand image generation.

When a user requests an image, the workflow triggers a sub-flow calling DALL-E 3 with the prompt and user ID, generates the image, and sends it back via Telegram. An additional Telegram node ensures stability by masking unsupported HTML syntax in AI responses. This setup handles multi-turn conversations seamlessly, making it ideal for interactive bots.

Benefits include time-saving automation of customer interactions, persistent memory for personalized engagement, and multimodal capabilities (text + images). In real estate lead management, use it to qualify leads via natural chat, generate property visuals from descriptions, schedule viewings, or provide market insights—boosting conversion rates without manual intervention. Highly customizable for any domain requiring conversational AI.