n8n AI Content Automation
Budget / Salary€250–750
TypeFreelance project
LocationRemote
Posted1 hour ago
I need a robust n8n workflow that orchestrates two multi-model AI agents—Claude and Gemini—to take a topic cue, run rapid web research, and return publish-ready social media posts with captions.
Here’s the flow I picture: a trigger (manual, schedule, or webhook) passes a prompt or keyword set into the workflow; Claude gathers and summarises the freshest, reputable information; Gemini then shapes that distilled insight into engaging, on-brand copy for Instagram, LinkedIn, and Twitter. The system should also be able to pivot to longer-form pieces—articles, reports, or summaries—when I choose, but high-velocity social content is the main output I’ll request day-to-day.
Key deliverables
• A working n8n .json workflow file with all nodes configured, including clear separation of Claude (research / summarisation) and Gemini (content generation) stages
• Environment variables or credentials placeholders for the two AI APIs, so I can drop in my keys without rewiring the flow
• Inline markdown or comments explaining each node’s purpose, plus a quick-start note describing how to trigger, test, and tweak tone or length
• A brief video or screenshot walkthrough demonstrating the workflow producing at least three sample social posts from a single prompt
Acceptance criteria
• The workflow handles research, content creation, and summary tasks end-to-end with no manual copy-paste steps
• Output: three platform-specific posts (text only is fine) generated in under two minutes from trigger
• All functions run inside n8n without external scripts other than the official Claude and Gemini endpoints
• Delivered file imports into the latest n8n release without errors
If you’ve already connected large-language models in n8n or built similar research-to-content automations, you’ll fly through this. Once I have the core flow, I’ll extend it to auto-schedule posts—so clean, well-commented nodes matter.
Here’s the flow I picture: a trigger (manual, schedule, or webhook) passes a prompt or keyword set into the workflow; Claude gathers and summarises the freshest, reputable information; Gemini then shapes that distilled insight into engaging, on-brand copy for Instagram, LinkedIn, and Twitter. The system should also be able to pivot to longer-form pieces—articles, reports, or summaries—when I choose, but high-velocity social content is the main output I’ll request day-to-day.
Key deliverables
• A working n8n .json workflow file with all nodes configured, including clear separation of Claude (research / summarisation) and Gemini (content generation) stages
• Environment variables or credentials placeholders for the two AI APIs, so I can drop in my keys without rewiring the flow
• Inline markdown or comments explaining each node’s purpose, plus a quick-start note describing how to trigger, test, and tweak tone or length
• A brief video or screenshot walkthrough demonstrating the workflow producing at least three sample social posts from a single prompt
Acceptance criteria
• The workflow handles research, content creation, and summary tasks end-to-end with no manual copy-paste steps
• Output: three platform-specific posts (text only is fine) generated in under two minutes from trigger
• All functions run inside n8n without external scripts other than the official Claude and Gemini endpoints
• Delivered file imports into the latest n8n release without errors
If you’ve already connected large-language models in n8n or built similar research-to-content automations, you’ll fly through this. Once I have the core flow, I’ll extend it to auto-schedule posts—so clean, well-commented nodes matter.
Apply on Freelancer →
Project sourced from Freelancer.com. Applications happen directly on the original platform — we never collect your data.