n8n AI Agent Workflow: Extract Tasks from Text with OpenRouter

Extract clear tasks, assignees, and deadlines from text—automatically—with an n8n AI Agent using OpenRouter

Have messy notes, emails, or messages that contain actionable items? This n8n AI Agent Workflow reads your input text, asks the best available OpenRouter chat model to extract tasks, assignees, and deadlines, and returns validated JSON—with built-in fallback retries if an API call fails.

What this workflow does

• Runs when you manually execute the workflow (manual trigger).
• Calls the OpenRouter Models API to fetch the current model catalog.
• Filters for free, text-capable chat models, then ranks candidates by factors like context length, estimated parameter size, vendor reputation, and supported features.
• Sends the extraction prompt plus your input text to the selected OpenRouter model via an n8n AI Agent.
• If the OpenRouter call fails (e.g., rate limits), it automatically retries with the next ranked fallback model until one succeeds or the list is exhausted.
• Parses the AI response into a clean JSON object containing a summary and structured action items.

Use cases

• Turn meeting notes into an actionable task list with owners and due dates.
• Extract project action items from project updates or support tickets.
• Convert informal “to-do” notes into structured JSON for downstream automation in your n8n workflows.
• Use a reliable OpenRouter fallback strategy to keep extraction running during provider issues.

Technical details

• Nodes / tools used: Code, Sticky Note, HTTP Request, Manual Trigger, n8n-nodes-langchainagent (AI Agent), n8n-nodes-langchainlm chat open router.
• Integration: OpenRouter API for model metadata and chat completions.
• Setup: Add an OpenRouter API key credential for the OpenRouter Chat Model connection, then update the prompt and the “Input Text to Process.”