Additional Features

Integrate with an AI agent

Feed an AI assistant accurate, current Invofox docs.

Building your integration with an AI assistant or coding agent? Three ways to feed it accurate, current Invofox docs — a ready-to-paste prompt, the llms.txt docs index, or the MCP server — plus a copy-paste AI context briefing at the end for product-level grounding.

Option A · Copy a prompt

No setup required. Copy the prompt below and paste it into ChatGPT, Claude, Gemini or your coding agent. It turns the assistant into a guided Invofox onboarding — it asks you what it needs (API key, document type, how you want results) and walks you through the integration one step at a time.

prompt
I want to integrate Invofox, a document extraction and parsing API that turns documents (invoices, receipts, bank statements, and more) into structured data. Act as my integration guide.
Use the official Invofox docs so your details are accurate — don't guess endpoints or invent field names. The basic ("happy path") flow is covered here:
- Integration guide (end-to-end happy path): https://developers.invofox.com/documentation/integrating-invofox
- Upload a document: https://developers.invofox.com/api-reference/ingest-reference/files/upload-direct-controller-upload-direct
- Get document information: https://developers.invofox.com/api-reference/main-reference/documents/get-document-info
- The document.processed webhook: https://developers.invofox.com/webhooks-reference/webhooks-reference/document-processing/receive-document-processed
(If you have the Invofox MCP server connected, you can also search the docs live.)
Guide me step by step, and BEFORE writing any code, ask me what you need instead of assuming. In particular, stop and ask me about:
- My API key — I may not have one yet; tell me where to create it if so.
- The document type — whether I already have a document-type ID, or want Invofox to auto-detect it.
- How I want to receive results: webhooks (recommended for production, needs a public endpoint) or polling (simplest to start with).
Keep it to the basic flow: upload a file, then get the extracted JSON back. Don't add Splitter, ZIP or the feedback loop unless I ask for them. Use whatever language or framework I'm already working in.
Go one step at a time, and wait for my answer before moving on.

Option B · Point it at llms.txt

Invofox publishes its documentation in the llms.txt format — a plain-text index built to be read by language models. Hand either URL to any assistant that can fetch a link and it pulls accurate, current docs with nothing to paste. No setup, and it never goes stale the way a hand-written prompt can.

Docs index — a curated map of the docs with links and short descriptions. Best when the agent can follow links.

url
https://developers.invofox.com/llms.txt

Full docs — the entire documentation concatenated into one markdown file. Best when you want to drop everything into the model’s context at once.

url
https://developers.invofox.com/llms-full.txt

Option C · MCP server

The Invofox MCP server exposes a searchDocs tool that lets any MCP-compatible agent query this documentation in real time. Works with Claude Code, Cursor, and any agent that supports the Model Context Protocol.

Claude Code:

Claude Code
$claude mcp add --transport http invofox https://developers.invofox.com/_mcp/server

Cursor, or manual setup — add to .mcp.json / .cursor/mcp.json:

.mcp.json
1{
2  "mcpServers": {
3    "invofox": {
4      "type": "http",
5      "url": "https://developers.invofox.com/_mcp/server"
6    }
7  }
8}

AI context

The llms.txt index (Option B) and the MCP server (Option C) already give an assistant the full technical reference — endpoints, schemas, concepts, webhooks. What they don’t cover is the product-level picture: how accurate Invofox is, how it’s priced, and how it handles security. This briefing fills exactly that gap, so it complements the docs rather than duplicating them. Paste it as background context.

context — paste into your assistant
# Invofox — product context for AI assistants
# For API endpoints, schemas and concepts, point the assistant at the docs index
# (https://developers.invofox.com/llms.txt) or the MCP server. This briefing covers
# only the product-level facts those technical docs leave out.
WHAT IT IS
Invofox is a document-intelligence (IDP) platform behind one REST API: it turns PDFs and images — invoices, receipts, delivery notes, purchase orders, bank statements, payslips, utility bills, ID documents, contracts, tax forms, US mortgage documents, plus custom document types — into clean, validated JSON. OCR + AI, no templates and no manual data entry.
INPUT — FORMATS & LIMITS
- Accepted formats: PDF, images (JPEG/JPG, PNG, TIFF), and ZIP archives bundling several files.
- Maximum file size: 100 MB per file (larger uploads are rejected with HTTP 413 Payload Too Large).
- Max pages per file and API rate limits are not publicly documented — confirm with Invofox if your volumes are high.
INGESTION CHANNELS
Documents can be sent three ways — the REST API, the web dashboard, and email forwarding. Every upload, whatever the channel, becomes an Import.
ACCURACY MODEL (the core promise)
- 99%+ accuracy backed by a per-field SLA.
- On enterprise plans, documents you send corrections on may not count toward your usage.
- ~99.2% average accuracy in production; ~0.4% of pages auto-flagged for human review.
SPEED & SCALE
- ~0.8s p50, under 2s average per page.
- 2M+ documents processed daily; handles large files (hundreds to thousands of pages) and traffic peaks.
LANGUAGES
English, Spanish, Portuguese, French, Italian, German and other Latin-script languages; other scripts on request.
SECURITY & COMPLIANCE
SOC 2 Type II, ISO 27001, GDPR, HIPAA. Encryption: TLS 1.3 in transit, AES-256 at rest with KMS-managed keys. Zero-retention mode (process, deliver, hard-delete — nothing kept). Data processed in EU or US regions; on-premises and VPC deployment available.
PRICING MODEL
Per page — no tokens or credits. Free tier to start; volume tiers and enterprise contracts for high throughput.
INDUSTRIES
ERP & accounting, mortgage & lending, insurance, healthcare, manufacturing, legal, payroll, shipping & logistics.