# [How C3] n8n - Workflow Automation n8n is a workflow automation platform that gives technical teams the flexibility of code with the speed of no-code. With 400+ integrations, native AI capabilities, and a fair-code license, n8n lets you build powerful automations while maintaining full control over your data and deployments. The ability to drive these flows via AI agents make it stand out. There are two ways to get started with n8n: Either you decide to use the n8n cloud thats accessible via https://n8n.io or you decide to self-host the [community edition](https://github.com/n8n-io/n8n) in your homelab or on a small VPS it doesnt't really take much resources. As n8n is pretty popular those who decide to self-host will find all sorts of [tutorials on youtube](https://www.youtube.com/results?search_query=how+to+install+n8n) which will help you to spin it up. I've installed the app on my unraid server but the cloud should look the same. ## Create a comput3 Agent Node Once installed you'll be greeted by the login screen. Login with your credentials. [<img src="https://hackmd.io/_uploads/ry2PbWBSxl.png"/>](https://hackmd.io/_uploads/ry2PbWBSxl.png) Create your first workflow. This is the no-code interface of n8n. It allows you to input content from whatever source (lets say a telegram bot) and route it through a node-based automation flow that can consists of hundreds of natively integrated services aswell as tens of thousand others as n8n has recently added mcp support. [<img src="https://hackmd.io/_uploads/ryDAWbHrxg.png"/>](https://hackmd.io/_uploads/ryDAWbHrxg.png) As the possibilities are endless we'll focus on setting up a comput3 based agent node in this tutorial as most of the AI driven n8n flows resolve around that particular node. Hit the plus button on the top right and search "agent" and select the "AI Agent" node. [<img src="https://hackmd.io/_uploads/ryglMbrBxl.png"/>](https://hackmd.io/_uploads/ryglMbrBxl.png) [<img src="https://hackmd.io/_uploads/ByLWzWrBee.png"/>](https://hackmd.io/_uploads/ByLWzWrBee.png) With the AI Agent node being added to your workflow its time to fuel it with a chat model. Click the plus button in the Chat Model connector below the AI Agent node and find the "OpenAI Chat Model" language model in the sidebar that pops up. We use that one as the comput3 llm api is built on the same standard, allowing for a low friction integration. [<img src="https://hackmd.io/_uploads/r1NGG-Brxg.png"/>](https://hackmd.io/_uploads/r1NGG-Brxg.png) In the view that pops up you now have to connect your comput3 account. In the connect dropdown hit "Create new credential" to start that process. [<img src="https://hackmd.io/_uploads/BkOWJMBHxx.png"/>](https://hackmd.io/_uploads/BkOWJMBHxx.png) In that view Copy + Paste your API Key and Base URL and hit save. In case your credentials are correct you'll receive a confirmation stating that the connection has been tested successfully. Change the name to comput3 in the top left corner and close the window. - Base URL: https://api.comput3.ai/v1 - API Key: your-comput3-api-key (scroll to the end of this post to find it) [<img src="https://hackmd.io/_uploads/HkmOB-SSxe.png"/>](https://hackmd.io/_uploads/HkmOB-SSxe.png) As this is originally an OpenAI connector n8n presets but once you click the the dropdown it'll fetch the available models from the api so you can change this to llama3:70b meaning that every workload thats going through this agent node is gonna be routed through gpus in the comput3 network. Rename it to "Comput3 Chat Model" and go back to canvas. [<img src="https://hackmd.io/_uploads/By7SrWBSxg.png"/>](https://hackmd.io/_uploads/By7SrWBSxg.png) [<img src="https://hackmd.io/_uploads/SkqOUZrBel.png"/>](https://hackmd.io/_uploads/SkqOUZrBel.png) You now have a blank canvas around your agent node. The settings that you saved for the comput3 models can be re-used along the way so every time you create a new agent node and add an "OpenAI Chat Model" you'll be able to select your comput3 credentials in the parameters and hit save - even in entirely new workflows. [<img src="https://hackmd.io/_uploads/B1M9IZSSxx.png"/>](https://hackmd.io/_uploads/B1M9IZSSxx.png) ## Start Automating your Processes Now you might ask what do I do with this? I really cannot answer that question for you as there are infinite possibilities. I recommend you to browse through and watch youtube videos to learn and understand what folks are automating with n8n. There are many step-by-step tutorials that resolve around the agent node that we've setup here. - Reddit https://www.reddit.com/r/n8n/ - Youtube https://www.youtube.com/results?search_query=best+n8n+usecases In addition to community inspiration and online tutorials there are many sources that allow you to download from Github that you can download for free. I highly recommend you to learn using n8n as its no-code nature and the huge community enables you to automate all sorts of things for your private and businesses needs. Everything fueled by comput3. - https://github.com/gyoridavid/ai_agents_az - https://github.com/wassupjay/n8n-free-templates - https://github.com/SaiAkhil066/n8n_50-50_challenge - https://github.com/Zie619/n8n-workflows There are many more resources. Drop the gud stuff on the Discord and I'll add them to the list so that everybody in the community can dig in and spread the word. # Setup your comput3 Account After you've [acquired](https://dexscreener.com/solana/34qhkhrhyningwruftqjnb2vfv8oqyqd5tvdwwde1man) and [staked](https://app.streamflow.finance/staking/solana/mainnet/DGZebyog1twdFaKwN2KZkBLxDkzX4rh7L61S6TAty8et) some $com token you can [login to your dashboard](https://launch.comput3.ai/) with your solana wallet and grab your API key at the bottom page. You can do way more things in this dashboard like spin up comfyui instances or launch dedicated ollama instances with broader model selection but for standard 70b llm inference via the default comput3 base url (https://api.comput3.ai/v1) this is not needed. In this tutorial we keep it simple. [<img src="https://hackmd.io/_uploads/BJGO16mreg.png"/>](https://hackmd.io/_uploads/BJGO16mreg.png) If you run into issues join the [official comput3 Discord server](https://discord.gg/EFZPGm9qUu) and ask the community.