Turn your Telegram bot into a smart front desk for support teams. It listens to new messages and turns audio, images, and text into clear replies and actions. Great for teams that work in Telegram and need quick triage.

An HTTPS webhook receives updates from Telegram. A gate checks user and chat IDs to allow only approved traffic. A router sends audio, text, or photos down the right path. Voice notes are downloaded and transcribed with OpenAI, then classified to decide if they are tasks. Text is scored by a separate OpenAI prompt to split into task or other. Photos are fetched, converted to base64, and analyzed with OpenAI vision to produce a short insight. Replies are sent back to Telegram with tailored messages. Built in tools set and verify webhooks for test and production.

You need a Telegram bot token, an OpenAI API key, and a public HTTPS URL for n8n. Expect faster intake, less manual reading, and fewer errors. Teams often cut sorting time from 30 minutes to 2 minutes and handle more chats without more staff. This fits frontline support, internal help desks, and field teams that share photos and voice notes from the job site.

• Webhook listener pulls Telegram updates into n8n in real time.
• User and chat validation blocks unknown senders before routing.
• Message router detects audio, text, or image and sends each to the right path.
• Audio flow downloads voice notes, transcribes with OpenAI, then classifies intent.
• Text flow normalizes fields and uses an OpenAI prompt to label task vs other.
• Image flow fetches the photo, converts to base64, and runs OpenAI vision analysis.
• Telegram reply nodes send tailored messages for task outcomes or general replies.
• Webhook tools set test and production URLs and check status with Telegram API.
• Error path sends a clear message when input is not allowed or cannot be processed.

• Reduce triage time from 30 minutes to 2 minutes per request
• Automate up to 80% of message classification across text, audio, and images
• Handle up to 5 times more chats with the same team size
• Improve response accuracy by 90% with consistent rules and prompts
• Connect Telegram and OpenAI without custom code or servers

1. Import the template into n8n: Create a new workflow in n8n > Click the three dots menu > Select 'Import from File' > Choose the downloaded JSON file.
2. You'll need accounts with Telegram and OpenAI. See the Tools Required section above for links to create accounts with these services.
3. Create a Telegram bot with BotFather and copy the bot token.
4. In n8n, open any Telegram node, click the Credential dropdown, choose Create new credential, and paste your bot token. Save the credential with a clear name.
5. In your OpenAI account, create an API key. In n8n, double click an OpenAI node, choose Create new credential, and paste the API key. Save the credential.
6. Open the Listen for Telegram Events node and set a unique path. Make sure your n8n instance has a public HTTPS URL.
7. Open the node named Telegram Token & Webhooks and fill in your bot token, your test URL, and your production URL for the webhook endpoint.
8. Run the Set Webhook Test URL node, then run Get Telegram Webhook Info to confirm the webhook is active for your test URL.
9. Open the Validation or Check User & Chat ID node and add your allowed user IDs and chat IDs so only approved senders are processed.
10. Send a text, a voice note, and a photo to your bot. Confirm that you receive the correct replies and that messages are labeled as task or other.
11. Tune the prompts in the gpt-4o-mini and gpt-4o-mini1 nodes to match your support rules. Save and test again.
12. When ready for live use, run Set Webhook Production URL and then Production Webhook Status to confirm the production webhook is active.
13. If the webhook does not work, make sure the URL is HTTPS and reachable, check that no other webhook is set for the bot, and verify your OpenAI quota is active.