Create & share AI videos with Telegram, Gemini & post to TikTok, Instagram, FB

Video Generation with Telegram Bot and Gemini API – Auto-Post to TikTok, Instagram & Facebook This n8n workflow turns your Telegram messenger into a full video content pipeline: you send a text or voice idea to a Telegram bot, collaborate with an AI on the script and caption, let Gemini generate the video, and then automatically publish it to TikTok, Instagram and Facebook – all with status tracking and Telegram confirmations. --- What You Need to Get Started This workflow connects several external services. You’ll need: Telegram Bot API Key Create a bot via Telegram’s BotFather and copy the bot token. This is used by the Listen for incoming events and other Telegram nodes. OpenAI API Key Required for: Speech to Text (OpenAI Whisper) – to transcribe voice notes. The OpenAI Chat Model used inside the AI Agent. Google Gemini / Vertex (via n8n Gemini Node) Used by the Generate a video node (model: veo-3.0-fast-generate-001) to create the video. Google Drive The generated video is temporarily stored in a Google Drive folder via Upload video and later downloaded again for uploading to Blotato. Blotato API Key Social media publishing layer: Uploads the video as media. Creates posts for TikTok, Instagram and Facebook. Exposes a status endpoint the workflow uses for polling. Google Sheets Access (optional but included) Used by Save Prompt & Post-Text to log your prompts and captions. Once these credentials are configured in n8n, the workflow runs end-to-end from Telegram idea to multi-platform publishing. --- How the Workflow Operates – Step by Step 1. Input & Initial Processing (Telegram + Voice Handling) This phase listens to your messages and normalizes them into a single text input for the AI. | Node Name | Role in Workflow | | ------------------------------ | ------------------------------------------------------------------------------------------------------- | | Listen for incoming events | Telegram Trigger node that starts the workflow whenever your bot receives a message (text or voice). | | Voice or Text | Set node that structures the incoming payload and prepares a unified text field for downstream nodes. | | A Voice? | IF node that checks whether the incoming message is a voice note or plain text. | | Get Voice File | If voice is detected, this Telegram node downloads the audio file from Telegram. | | Speech to Text | Uses OpenAI Whisper to convert the voice note into a text transcript. | If you send text: The workflow skips the voice download/transcription and goes directly to the AI agent with your original text. If you send a voice note: The workflow downloads the file, runs it through Whisper in Speech to Text, and passes the resulting transcript onward. The output of this stage is always a clean text string representing your idea. --- 2. AI Core & Approval Loop (Script + Caption via AI) Here the AI designs the video prompt (for Gemini) and the social media caption (for all platforms), then iterates with you until you approve. | Node Name | Role in Workflow | | -------------------------------------- | -------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | AI Agent | Central logic agent. Takes your idea text and applies the system prompt to create a video prompt and social media caption. Handles revisions based on your feedback. | | OpenAI Chat Model | The LLM backing the agent (OpenAI Chat model). | | Window Buffer Memory | Memory buffer that stores recent turns, so the agent can keep context across your “make it shorter / more fun / more technical” requests. | | Send questions or proposal to user | Telegram node that sends the AI’s suggested prompt + caption back to you for review. | | Approved from user? | IF node that checks whether the agent’s output is the “approved” JSON (meaning you said “ok” / “approved”) or just a normal suggestion. | Video Prompt Assistant System Prompt (Internal Instructions) Th