anaksor-ocr
Sokheng · Kit
Reads Khmer and English text from documents, receipts and photos of signs, using a detector for text lines and a recogniser for each line. Suits developers who…
An n8n workflow that runs a small Telegram chatbot with three behaviours. It chats in the user’s language, welcomes new users and draws a picture when a message starts with an image command.
Telegram chatbot with a welcome and an image command is an n8n workflow in the Productivity category, listed by Mindket. It is free. This is version 1, last updated on . Download it on this page or copy its JSON, then import it into an n8n canvas. Nothing runs until you activate it in n8n.
A Telegram message
It starts when someone sends your bot a message:
A reply in the Telegram chat: a chat answer, a welcome, a photo or a help text
What you need before you start.
Chat
Takes each message, shows the typing sign and sends the replies and the photo.
NeedsA bot token from BotFather, in a Telegram credential
Chat model and picture model
Writes the chat replies and the welcome with gpt-5.6-terra and draws pictures with gpt-image-2.6-flare.
NeedsAn OpenAI API key, in an OpenAI credential
Checked by Mindket on 13 Nov 2026
This n8n workflow runs a small Telegram chatbot on OpenAI. A message that does not start with a slash gets a chat reply in the user’s own language, with a few emojis. The command /start gets a welcome message. The command /image, followed by a request, draws a picture and sends it as a photo. Any other command gets a short help text. Twelve nodes do the work, beside four sticky notes.
Beginners who want a first Telegram bot with a few commands, and small teams who want a friendly assistant that replies in the language the user writes. Skip it if you need memory, because each message is answered on its own.
| Industry | Who uses it | How they use it | Example |
|---|---|---|---|
| Community | Community managers | Greet new members and answer their questions in their language | A manager’s bot welcomes a Spanish speaker in Spanish and answers a question. |
| Education | Teachers | Let students try an AI assistant and ask it for pictures | A teacher shows a class how to ask the bot for a drawing of a volcano. |
| Small business | Shop owners | Offer a friendly chat helper next to the shop | An owner points customers to the bot for quick questions. |
| Learning n8n | Beginners | Study a switch with several routes | A beginner changes the welcome text and sees which route runs. |
You need basic n8n skills and two accounts: Telegram and OpenAI. Telegram sends each message to your n8n over HTTPS, so n8n needs a public address.
Any n8n that Telegram can reach over HTTPS and that can reach OpenAI will run it.
Download saves a zip with workflow.json and Mindket’s setup notes in your downloads folder. Two accounts are involved.
| Need | Detail |
|---|---|
| n8n | On n8n Cloud or on a server of your own with a public HTTPS address |
| Telegram | A bot made with BotFather, and its token in a Telegram credential |
| OpenAI | An API key in an OpenAI credential, used by the chat, the welcome and the picture nodes |
workflow.json.Telegram Trigger, Send Typing action, Text reply, Send image and Send error message.Chat_mode, Greeting and Create an image.gpt-5.6-terra, and the pictures use gpt-image-2.6-flare.Settings and read the system command. It sets the bot’s tone./start, a plain sentence and /image a red kite.Write to the bot in Telegram. A plain sentence gets a chat reply, the command /start gets the welcome, and a message that begins with /image and a space draws a picture. The bot shows a typing sign or an upload sign while it works.
| To change | Where |
|---|---|
| The bot’s tone and language rules | system_command in Settings |
| The chat and welcome model | modelId in Chat_mode and Greeting |
| The picture model | modelId in Create an image |
| The help text | The text in Send error message |
| The commands | The rules in CheckCommand |
A user sends the message /image a watercolour lighthouse at dusk.
The switch sends it to Create an image, which draws a lighthouse from the words after the command. Send image posts the picture in the chat as a photo.
If you are new. n8n draws a workflow as boxes joined by lines, and each box, a node, does one job. A switch is a box that sends each message down one of several lines. Here the first characters of the message pick the line.
For experts. The router is a Switch node, version 1, with three rules and a fallback output. A Merge node waits for the typing sign before the message is routed.
| Node | What it does | What to know |
|---|---|---|
PreProcessing | Makes sure message.text exists | A photo or sticker becomes an empty text, which the first rule sends to the chat step |
Settings | Holds the system command and the typing sign | The typing sign is upload_photo for /image and typing for the rest |
Merge | Waits for the typing sign and the settings | It passes on the settings data |
CheckCommand | Routes by the start of the text | Text with no slash, /start, /image with a space, and everything else |
Send image | Sends the picture from the step before as a photo | It reads the binary field data |
The old OpenAI node was replaced, and Mindket did not run the new chat or picture steps.
gpt-4, which OpenAI’s deprecations page, read on 12 November 2026, lists as shut down on 23 October 2026. The welcome step had no model, so the old node used its default gpt-3.5-turbo, shut down on the same date. Mindket’s copy uses gpt-5.6-terra for bothgpt-image-2.6-flare, which returns a file, so Send image now uploads a file_ or *, and a message holds at most 4096 characters (Telegram’s Bot API page)An n8n workflow that runs a small Telegram chatbot with three behaviours. It chats in the user's language, welcomes new users and draws a picture when a message starts with an image command.
workflow.json: the workflow as n8n exports it, sixteen nodes. Twelve do the work and four are sticky notes that label the routes. It holds no credential.README.md: these notes.gpt-4, which OpenAI's deprecations page, read on 12 November 2026, lists as shut down on 23 October 2026. The welcome step had no model, so the old node used its default, gpt-3.5-turbo, shut down on the same date. Both now use gpt-5.6-terra.gpt-image-2.6-flare, a model OpenAI's page lists for image generation.Send image used to send a link. It now sends the file that Create an image produces.Settings that held them. The model page lists no temperature setting, and reasoning tokens count against a token limit.workflow.json.Telegram Trigger, Send Typing action, Text reply, Send image and Send error message.Chat_mode, Greeting and Create an image.gpt-5.6-terra, and the pictures use gpt-image-2.6-flare.Settings and read the system command. It sets the bot's tone and asks for a reply in the user's language./start, a plain sentence and /image a red kite.Telegram Trigger starts the run for each update. PreProcessing makes sure message.text exists, and Settings adds the system command and the typing sign.Send Typing action shows typing, or upload_photo for an image command. Merge waits for it and the settings before the message goes on.CheckCommand routes by the start of the text. Text with no slash goes to Chat_mode. The text /start goes to Greeting. A text that starts with /image and a space goes to Create an image. Everything else goes to Send error message.Chat_mode and Greeting ask gpt-5.6-terra for a reply, and Text reply sends it in Telegram's Markdown mode.Create an image draws a picture from the words after the command, and Send image posts it as a photo.system_command in Settings.modelId in Chat_mode and Greeting.modelId in Create an image.Send error message.CheckCommand.message.content, which is where Text reply reads it.PreProcessing turns it into an empty text, and the first rule of CheckCommand sends it to the chat step as an empty message._ or *, and a message holds at most 4096 characters (Telegram's Bot API page).NOTICE.md 549 B README.md 5 KB workflow.json 10 KB
Sokheng · Kit
Reads Khmer and English text from documents, receipts and photos of signs, using a detector for text lines and a recogniser for each line. Suits developers who…
Qwen · Kit
Qwen2.5-7B-Instruct as GGUF files in eight quantization sizes, to chat with locally through llama.cpp. For people who want a capable 7.6B model on their own co…
ACE-Step · Kit
Generates music from text with a 4B diffusion decoder, 50 steps and guidance control, aiming for higher audio quality than the base. For creators with a GPU of…
op7418 · Skill
Skill that makes single-file HTML slide decks, slide illustrations and platform covers in two visual styles, with a presenter mode for rehearsal. For speakers…
zarazhangrui · Skill
Builds HTML presentations from scratch or from a PowerPoint file and shows visual style previews to pick from, for people who want web slides with no CSS or Ja…
iptv-org · Kit
Collects publicly available live TV channels from many countries into playlists that any video player with live streaming can open. Made for viewers and for de…