Get Started
Back to Templates

Telegram Bot: Analyze Images with GPT-4o-Mini/NVIDIA Vila & Generate Images with Stable Diffusion 3

Created by

Created by: Cheng Siong Chin || cschin

Cheng Siong Chin

Last update

Last update 2 days ago

Categories

Share

Introduction

Transform your Telegram bot into an AI vision system using GPT-4o-Mini and NVIDIA Stable Diffusion 3. Perfect for content moderators, researchers, and developers.

Workflow Explanatory

  1. At start: Processes Telegram messages: images→analysis, text→image generation
  2. At Router: Routes by content type
  3. Upper path: Analyzes images using Nvidia Vila + GPT-4o-Mini
  4. Lower path: Generates images from text via Stable Diffusion 3
  5. At Merge: Combines AI results
  6. At Gmail: Emails processed results

How It Works

  1. Telegram Trigger listens for messages (images, text, documents)
  2. Content Router directs images → AI analysis, text → image generation
  3. Image Analysis: Downloads image → GPT-4o-Mini vision analysis → Email results
  4. Image Generation: Text prompt → Stable Diffusion 3 → Email generated image
  5. Gmail Notifications send formatted reports

Prerequisites

  • Telegram Bot token (via @BotFather)
  • OpenAI API key (GPT-4 Vision)
  • NVIDIA API key (free tier available)
  • Gmail OAuth2 credentials

Setup Steps

Setup Steps

  1. ** Create Telegram Bot** - Create Telegram bot and obtain token
  2. ** Configure API Credentials** - Configure API credentials in HTTP Request nodes
  3. ** Set Up Gmail OAuth2** - Set up Gmail OAuth2
  4. ** Import and Activate Workflow** - Import workflow, update credentials, and activate

Customization Options

  • Add more AI models (Anthropic, Gemini)
  • Route audio/documents to transcription/OCR
  • Replace Gmail with Slack or Discord
  • Connect to databases for storage

Benefits

  • Speed: Seconds per analysis vs. hours manually
  • Accuracy: AI-powered visual understanding
  • Intelligence: Historical tracking enables trend analysis