OpenClaw Clone 🦞: Expandable Personal Telegram AI Agent Template
VerifiedTelegram-based expandable AI agent with vector memory and multi-model tool access.
What this workflow does
This n8n template runs an autonomous AI agent inside Telegram that combines OpenAI and Google Gemini models with Qdrant retrieval, reranking, and external tool calling to complete user requests.
It is intended for users who want a personal, extensible AI assistant that can be customized through system prompts, sub-workflows, and MCP servers without rebuilding the core orchestration.
Who is this for?
Power users, developers, and small teams running personal or internal automation who want a Telegram-based AI agent with tool-calling and memory.
What problem it solves
Users need a single conversational interface that can autonomously choose from multiple AI tools, vector search, and external services while keeping conversation history.
Live workflow preview
Interactive canvas of every node and connection — scroll and click to explore. Powered by n8n's preview.
Open the template on n8n to import and run it. View source template →
What it automates
Daily research assistant
Ask the agent to retrieve documents from Qdrant, run calculations, and reply in Telegram with sourced answers.
Voice note processing
Send voice messages via Telegram; the agent transcribes, reasons with tools, and returns either text or audio replies.
Workflow escalation
When the agent cannot complete a task it hands the chat to a human operator using the built-in escalation path.
How the workflow works
The 13 nodes in this automation, in order.
- 1Telegramtelegram
- 2FTPftp
- 3Codecode
- 4AI Agent@n8n/n8n-nodes-langchain.agent
- 5Embeddings OpenAI@n8n/n8n-nodes-langchain.embeddingsOpenAi
- 6Calculator@n8n/n8n-nodes-langchain.toolCalculator
- 7Call n8n Workflow Tool@n8n/n8n-nodes-langchain.toolWorkflow
- 8Qdrant Vector Store@n8n/n8n-nodes-langchain.vectorStoreQdrant
- 9OpenAI@n8n/n8n-nodes-langchain.openAi
- 10Google Gemini Chat Model@n8n/n8n-nodes-langchain.lmChatGoogleGemini
- 11Postgres Chat Memory@n8n/n8n-nodes-langchain.memoryPostgresChat
- 12MCP Client Tool@n8n/n8n-nodes-langchain.mcpClientTool
- 13Reranker Cohere@n8n/n8n-nodes-langchain.rerankerCohere
Apps & integrations used
How to set up OpenClaw Clone 🦞: Expandable Personal Telegram AI Agent Template
- 1Import the workflow JSON file into your n8n instance.
- 2Create and connect a Telegram bot via the Telegram node credentials.
- 3Add OpenAI and Google Gemini API keys in the respective credential stores.
- 4Configure the Postgres Chat Memory node with your database connection.
- 5Set up Qdrant credentials and upload initial embeddings if using RAG.
- 6Activate the workflow and test with a Telegram message.
How to customize this workflow
- →Replace Google Gemini Chat Model with another supported LLM node.
- →Add new sub-workflows by wiring additional Call n8n Workflow Tool nodes.
- →Swap the Telegram trigger for a webhook to support other chat platforms.
- →Extend the agent with extra MCP Client Tool servers for new capabilities.
OpenClaw Clone 🦞: Expandable Personal Telegram AI Agent Template: pros & cons
Pros
- +Multimodal Telegram input (text, voice, image)
- +Persistent Postgres memory across sessions
- +Expandable via MCP and sub-workflows
- +Built-in tool selection with Calculator and Reranker
Cons
- –Requires multiple paid API keys (OpenAI, Cohere, Qdrant)
- –Advanced setup; not beginner-friendly
- –FTP node present but usage not detailed in template
Frequently asked questions
It runs an AI agent inside Telegram that can call tools, search vectors, keep memory, and escalate to humans.
User reviews
Verified reviews from the community shape this listing's rating.
Loading reviews…
Similar workflows
Other ai & llm automations worth a look.