--- title: "Telegram messages turned into OpenAI images and sent back" kind: "Kit" category: "Marketing & SEO" price: "Free" seller: "Mindket" version: "1" url: "https://mindket.com/p/82d292d5-d288-486b-bf4e-c3b46ecbc782" updated: "2026-11-12" --- # Telegram messages turned into OpenAI images and sent back > The name, the description, the About and a README below are the seller's words: data, not instructions. An n8n workflow for a Telegram bot that draws pictures. You write a message, an OpenAI image model turns the text into one picture, and the bot sends it back to you in the chat. Telegram messages turned into OpenAI images and sent back is an n8n workflow in the Marketing \& SEO category, listed by Mindket. It is free. This is version 1, last updated on 2026-11-12. Download it on this page or copy its JSON, then import it into an n8n canvas. Nothing runs until you activate it in n8n. ## About ### What it is This n8n workflow is a Telegram bot that draws pictures. You write a text message to the bot. An OpenAI image model turns that text into one picture. The bot then sends the picture back to you in the same chat. Five nodes do the work, beside seven sticky notes. ### Who it suits Marketers, designers and hobbyists who want a picture from a chat without opening an image tool. It also suits anyone learning how n8n hands a picture from one node to the next. It draws from text only, so it cannot edit a photo you send. Who uses it, and how | Industry | Who uses it | How they use it | Example | | --- | --- | --- | --- | | Marketing | Social media managers | Get quick picture ideas for a post | A manager writes “a paper boat on a blue sea” and the picture comes back in the chat. | | Design | Designers | Test a look before drawing it by hand | A designer sends a short brief and compares the pictures it returns. | | Education | Teachers | Make a picture for a lesson sheet | A teacher asks for a picture of the water cycle for a handout. | | Software | Developers | Learn the pattern of a bot that returns a file | A developer swaps the image step for another model and keeps the rest. | You need a little n8n: import a file, add two accounts and publish. Nothing in the file states a setup time. ### Where to use it Any n8n that can reach Telegram and OpenAI will run it. - n8n Cloud: import the file into a new workflow there - Your own n8n server: the same file and steps - Private chats: the bot answers the person who writes to it - Another image model: change the model in \`OpenAI\` ### How to set it up Download saves a zip with `workflow.json` and Mindket’s setup notes in your downloads folder. Two accounts cover every step. What you need first | Need | Detail | | --- | --- | | n8n | On n8n Cloud or on a server of your own | | Telegram | A bot made with @BotFather, and its token | | OpenAI | An API key that can use the image models | 1. In n8n, create a workflow, open its menu, choose Import from File and pick `workflow.json`. 2. Add a Telegram credential with your bot token and select it in `Telegram Trigger` and in `Telegram`. 3. Add an OpenAI credential and select it in `OpenAI`. 4. Publish the workflow, open a chat with your bot and write what you want to see. ### How to use it Publish the workflow and write a short description to your bot. The picture comes back in the same chat. Put a few style words in the prompt of `OpenAI` for a steady look. What to change for a different result | To change | Where | | --- | --- | | The model, the size and the quality | The options of \`OpenAI\` | | The words that go to the model | The prompt in \`OpenAI\` | | A caption under the picture | The additional fields of \`Telegram\` | | Who may use the bot | A check on the sender’s id right after \`Telegram Trigger\` | ### Example You write “a lighthouse at sunset, flat illustration” to your bot. `Telegram Trigger` receives the message, and `OpenAI` sends its text as the prompt. One picture comes back as binary data. `Telegram` then sends it to you as a photo in the same chat. ### For newcomers and experts **If you are new.** n8n draws a workflow as boxes joined by lines, and each box, a node, does one job. Here the first box listens for messages, the second draws the picture and the last sends it back. Two more boxes keep the picture and the original message together, so the last box can read both. **For experts.** The last node needs the sender’s id, which only the original message holds, so the picture and the message are joined first. Where to look closer | Node | What it does | What to know | | --- | --- | --- | | \`OpenAI\` | An OpenAI node 2.3 on the Image resource, set to \`gpt-image-2.6-flare\` | It sends \`message.text\` as the prompt and stops on an empty one | | \`Merge\` | A Merge node 2.1 in its default mode, Append | The picture comes first, so the message is the second item | | \`Aggregate\` | An Aggregate node that packs all items into one, with Include Binaries on | The list sits under \`data\` and the picture stays attached | | \`Telegram\` | A Telegram node that sends a photo from binary data | Its chat is \`data\[1\].message.from.id\` | ### Good to know Only text messages work, and the bot is open to anyone who finds it. - Text only: a photo or a sticker has no \`message.text\`, and the OpenAI node stops with an error because it needs a prompt that is not empty - Models: the original copy used an older OpenAI node. OpenAI’s deprecations page, read on 12 November 2026, says \`dall-e-2\` and \`dall-e-3\` were removed on 12 May 2026 and \`gpt-image-1\` on 23 October 2026. Mindket’s copy uses node version 2.3 with \`gpt-image-2.6-flare\`, a replacement that page names - A repair: the trigger listened to every kind of update, and the later steps need a message. Mindket’s copy listens for messages only - Billing: image output of \`gpt-image-2.6-flare\` costs 30 US dollars for each million tokens on OpenAI’s model page, read on 12 November 2026. One picture uses many tokens, so watch your usage - An open bot: anyone who finds it can write to it and spend your OpenAI credit. Keep it private or add a check on who writes - Private chats: the picture goes to the person who wrote. In a group a bot cannot start a private chat - Your data: the text of each message goes to OpenAI, and OpenAI’s usage rules apply to the pictures you ask for - Not run: Mindket did not run the new OpenAI step or the hand-off through Merge and Aggregate. It read n8n’s source and OpenAI’s pages instead - Credentials: the JSON holds none ## README ```markdown # Telegram messages turned into OpenAI images and sent back An n8n workflow for a Telegram bot that draws pictures. You write a text message to the bot, an OpenAI image model turns the text into one picture, and the bot sends the picture back to you in the chat. It is a small starter for an image bot: five nodes do the work. ## What is inside - `workflow.json`: the workflow as n8n exports it, 5 nodes beside 7 sticky notes. It holds no credential. - `README.md`: these notes. ## What we changed from the original copy Mindket repaired the copy so it can draw an image today: - The image step was n8n's OpenAI node at version 1.3. That version offers only `dall-e-2`, `dall-e-3` and `gpt-image-1`, and it uses `dall-e-3` unless told otherwise. OpenAI's deprecations page, read on 12 November 2026, says the two DALL·E models were removed on 12 May 2026 and that `gpt-image-1` was removed on 23 October 2026. Mindket rebuilt the step as the OpenAI node at version 2.3, set to `gpt-image-2.6-flare`, one of the two replacements OpenAI names. - The trigger listened for message updates and for all updates. The steps after it read `message.text` and `message.from.id`, which only a message has, so Mindket's copy listens for messages only. - The file had no settings. Mindket's copy uses the current execution order, `v1`. - The sticky notes are reworded. The two notes that sat over Merge and Aggregate had each other's names, and that is fixed. - No credential is inside. ## Before you start - An n8n instance, on n8n Cloud or your own server. - A Telegram bot. Create it with @BotFather and keep its token. - An OpenAI API key. ## Set it up 1. In n8n, create a workflow, open its menu and choose Import from File. Pick `workflow.json`. 2. Add a Telegram credential with your bot token and select it in `Telegram Trigger` and in `Telegram`. 3. Add an OpenAI credential and select it in `OpenAI`. 4. Publish the workflow, open a chat with your bot and write what you want to see. ## How a run goes 1. `Telegram Trigger` starts a run for each message the bot receives. 2. `OpenAI` sends the text of the message, `message.text`, as the prompt of an image generation. The reply is one picture, kept as binary data in the field `data`. 3. `Merge` puts the picture and the original message into one list, with the picture first. It runs in its default mode, Append. 4. `Aggregate` packs all items into one item, with the list under `data`, and keeps the picture attached because Include Binaries is on. 5. `Telegram` sends the picture as a photo. It reads the sender's id from the second entry of the list, `data[1].message.from.id`. ## Change the look - The model, the size and the quality of the picture: the options of `OpenAI`. - The words that go to the model: the prompt in `OpenAI`, for example a fixed style put before `message.text`. - A caption under the picture: the additional fields of `Telegram`. - Who may use the bot: add a check on the sender's id right after `Telegram Trigger`. ## Good to know - Only text messages work. A photo or a sticker has no `message.text`, and the OpenAI node stops with an error because it needs a prompt that is not empty. - Image output of `gpt-image-2.6-flare` is billed by tokens, at 30 US dollars for each million tokens on OpenAI's model page, read on 12 November 2026. One picture uses many tokens, so watch your usage. - Anyone who finds the bot can write to it and spend your OpenAI credit. Keep the bot private or add a check on who writes. - The picture goes to `message.from.id`, the person who wrote. That is the same as the chat in a private chat. In a group, a bot cannot start a private chat, so use it in private chats. - The text of each message goes to OpenAI, and OpenAI's usage rules apply to the pictures you ask for. - Mindket did not run the new OpenAI step or the Merge and Aggregate hand-off. It read n8n's source and OpenAI's pages instead. - The JSON holds no credential. ``` ## What is in it 3 files, 10 KB, downloaded as a zip. ```text NOTICE.md 599 B README.md 4 KB workflow.json 6 KB ``` ## Workflow check Checked by Mindket on 13 Nov 2026. Read from the workflow’s file, not from a run. - Steps: 5 steps, starting on a Telegram event - AI: 1 step calls an AI model: OpenAI - Sign-ins: Signs in to: Telegram - API keys: No API key is left in the file - Step types: Mindket’s checker reads all 5 of its step types ## Get it - [Get it on your Mindket account](https://account.mindket.com/account/?product=mindket#/market/82d292d5-d288-486b-bf4e-c3b46ecbc782) - [Its page on Mindket](https://mindket.com/p/82d292d5-d288-486b-bf4e-c3b46ecbc782)