Extract structured invoice JSON from PDFs with Mistral OCR and an LLM API

by Venkata V ✓ · 297 views · created 2026-04-22 · Invoice Processing, AI Summarization

What this workflow does

This workflow has 8 nodes and runs when an HTTP POST request hits its webhook URL. It uses only n8n's built-in nodes. It also calls api.mistral.ai over HTTP.

How it runs

Trigger: When an HTTP POST request hits its webhook URL.

1. Mistral OCR (Mistral AI): uses Mistral AI
2. Normalize OCR Text (Code): runs custom JavaScript code
3. LLM Extract JSON (HTTP Request): calls api.mistral.ai (POST)
4. Clean JSON (Code): runs custom JavaScript code
5. Confidence >= 0.5? (If): checks a condition and branches into true / false
6. Respond OK (Respond to Webhook): sends the response back to the webhook caller
7. Respond Review (Respond to Webhook): sends the response back to the webhook caller

HTTP calls to: api.mistral.ai

Before you upgrade to n8n 3.0

Nodes used (6)

• Webhookn8n-nodes-base.webhook
• Mistral AIn8n-nodes-base.mistralAi
• Coden8n-nodes-base.code
• HTTP Requestn8n-nodes-base.httpRequest
• Ifn8n-nodes-base.if
• Respond to Webhookn8n-nodes-base.respondToWebhook

From the author

N8N AI LLM Unstructured Invoice data PDF OCR recognition to JSON output API What this workflow does 1. Accepts a PDF or image upload via Webhook as binary property "data" 2. Runs OCR with the Mistral OCR node 3. Normalizes OCR text 4. Sends OCR text to an LLM…

View the original on n8n.io → Check your own copy