Automate Telegram Image Replies with OpenAI and n8n

Opening Problem Statement

Meet Sarah, a community manager for a popular Telegram group centered around art and creativity. Every day, Sarah spends hours manually responding to member messages with relevant images, links, or explanations to keep her audience engaged. This repetitive task not only consumes Sarah’s valuable time but also limits her ability to grow the community creatively. Miscommunication occasionally happens too, as replies may come late or lack personalization. Sarah needs a smart and automated way to reply with AI-generated images instantly to improve engagement and save countless hours.

What This Automation Does

This n8n workflow transforms incoming Telegram messages into AI-generated images and sends these images back to the user automatically. Here’s what happens when it runs:

• It triggers on any new message received via a Telegram bot.
• It sends the message text to OpenAI’s image generation API to create a relevant picture.
• It merges and aggregates the image data properly for handling within n8n.
• It sends the generated image back to the original Telegram user seamlessly.
• Processes messages in real-time for instant reply ensuring active engagement.
• Ensures binary image data is handled correctly to maintain quality and delivery integrity.

This automation can save Sarah hours daily by eliminating manual image searching and messaging, enhancing user satisfaction with smart AI responses.

Prerequisites ⚙️

• 🔑 Telegram Bot Account with API credentials set up to allow n8n to receive messages and send replies.
• 🔑 OpenAI API Key with access to image generation endpoints.
• 🔌 n8n Workflow Automation Platform – either cloud-hosted or self-hosted. For self-hosting options, consider Hostinger’s guide.

Step-by-Step Guide

1. Set up Telegram Trigger

Navigate to the Telegram Trigger node. Under the “Updates” field, select “message” and “*” to capture all new text messages from users. Connect your Telegram bot credentials to allow the webhook. This node will start the workflow every time a new message arrives.

Visual: You should see a webhook URL generated after credential setup, ready to receive Telegram updates.

Common mistake: Forgetting to link correct Telegram bot credentials or not enabling necessary Telegram bot permissions.

2. Configure OpenAI Node for Image Generation

Click on the OpenAI node (specifically the LangChain OpenAI node). In the prompt field, enter the expression: = {{$json.message.text}}. This passes the Telegram user’s message as input text to OpenAI’s image generation request.

Set the resource parameter to “image” to indicate that you want an image output. Connect your OpenAI API credentials here.

Visual: You’ll see the prompt field populated with the Telegram message text mapping.

Common mistake: Not setting the resource to “image” will cause text completion instead of image generation.

3. Use Merge Node to Organize Incoming Data

After the OpenAI node, add the Merge node. This node consolidates data from multiple branches—in this case, the Telegram trigger and OpenAI outputs—so that you can handle the combined dataset downstream.

Visual: The node will show combined inputs from the preceding nodes.

Common mistake: Not connecting the Telegram Trigger’s second output to the Merge node to maintain user info for sending replies.

4. Aggregate Node to Handle Binary Image Data

Add the Aggregate node next, configuring it to aggregate all item data including binaries like images. Enable the “includeBinaries” option to ensure image data is preserved.

Visual: Aggregated data node showing combined binary and JSON data.

Common mistake: Missing the “includeBinaries” option will lose image data during workflow processing.

5. Configure the Telegram Node to Send Photos

Finally, set up the Telegram node to send the AI image back to the user. Under “Operation,” pick “sendPhoto.” Use the expression = {{$json.data[1].message.from.id}} for “Chat ID” to correctly direct the reply back to the sender.

Select “binaryData” option to indicate you will send an image file as binary data.

Visual: The node preview should show intended photo sending with user chat ID dynamically set.

Common mistake: Using static chat IDs or forgetting to enable binary data sending.

6. Activate the Workflow and Test

Save and activate your workflow. Test it by sending a message to your Telegram bot. The bot should reply instantly with an AI-generated image matching your message content.

Expected: Fast, accurate, and visually relevant images delivered back automatically.

Customizations ✏️

• Customize AI Prompt: Change the OpenAI node’s prompt to add context or improve image detail, like prefixing with “Draw a beautiful…” for artistic outputs.
• Message Filters: Add a filter node before OpenAI node to only process messages with specific keywords.
• Send Alternate Media Types: Modify the Telegram node to send videos or documents by changing the “Operation” and handling appropriate binary data.
• Reply with Captions: Use added parameters in Telegram node’s “Additional Fields” to send captions with images.
• Logging User Interactions: Add a database node or Google Sheets node to save user messages and generated image links for tracking.

Troubleshooting 🔧

Problem: “No image generated or empty response from OpenAI”

Cause: Incorrect prompt input or API resource misconfigured.

Solution: Double-check that the prompt in the OpenAI node uses = {{$json.message.text}} and that the resource is set to “image.”

Problem: “Telegram message not sent or wrong chat ID”

Cause: Chat ID is not dynamically set or duplicated improperly.

Solution: Ensure the Telegram node’s chat ID is correctly set to = {{$json.data[1].message.from.id}} from merged inputs, not hardcoded.

Pre-Production Checklist ✅

• Confirm Telegram bot credentials have necessary permissions and bot is active.
• Verify OpenAI API key is valid and has image generation enabled.
• Test triggering the webhook by sending sample Telegram messages.
• Check that the Merge node combines data as expected.
• Ensure image binary data is correctly included and sent by the Telegram node.

Deployment Guide

After testing thoroughly, activate the workflow in n8n. Keep monitoring execution logs for any errors. For scale, consider setting up alerts on workflow failures and maintaining your OpenAI API usage within limits.

FAQs

• Can I use this workflow with WhatsApp instead of Telegram?
  Currently, this workflow explicitly uses the Telegram Trigger and Telegram nodes; adapting it to WhatsApp would require swapping those components with WhatsApp-compatible nodes.
• Does this consume many OpenAI API credits?
  Image generation is costlier than text responses; monitor usage carefully to avoid unexpected costs.
• Is user data secure?
  n8n stores data temporarily during execution; ensure you have proper security and compliance according to your deployment environment.

Conclusion

By setting up this AI-powered Telegram image reply automation, you, like Sarah, can reclaim hours spent on repetitive tasks, delight community members with instant creative responses, and scale up your engagement effortlessly. This workflow showed you how to connect Telegram with OpenAI through n8n to automate image requests and responses seamlessly. Next, consider adding personalized image captions or tracking user interactions for deeper insights. Embrace AI today to elevate your Telegram community interaction!