DocMind API

AI Document Intelligence API -- parse receipts, invoices, bank statements, contracts, and any document into structured JSON using your choice of LLM provider.

Live Documentation | GitHub

---

Features

• Multi-LLM support -- route requests through Anthropic Claude, OpenAI GPT, or Google Gemini
• Five document types -- receipts, invoices, bank statements, contracts, and general documents
• Flexible input -- file upload (multipart), base64-encoded images, or raw text
• Batch processing -- parse up to 20 documents in a single request
• Conversational AI -- chat endpoint for general-purpose LLM access
• API key management -- tiered access control (free / pro / enterprise) with built-in rate limiting
• Zero external databases -- runs entirely in-memory; swap in your own persistence layer as needed

---

Quick Start

# Clone the repository
git clone https://github.com/sjgant80-hub/docmind-api.git
cd docmind-api

# Install dependencies
npm install

# Copy the example env file and add your keys
cp .env.example .env
# Edit .env with your LLM provider key(s)

# Start the server
npm start

The server starts on http://localhost:3000 by default. A demo API key is printed to the console on startup.

For development with auto-reload:

npm run dev

Requirements: Node.js >= 18

---

Environment Variables

Variable	Description	Default
PORT	Server port	3000
NODE_ENV	Environment (production / development)	--
ADMIN_KEY	Master key for admin endpoints	Auto-generated
ANTHROPIC_API_KEY	Anthropic API key (Claude)	--
OPENAI_API_KEY	OpenAI API key (GPT)	--
GOOGLE_API_KEY	Google AI API key (Gemini)	--
DEFAULT_PROVIDER	Default LLM provider: anthropic, openai, or google	anthropic
RATE_LIMIT_PER_MINUTE	Global rate limit (requests per minute)	60
USAGE_WEBHOOK_URL	Optional webhook URL for usage event notifications	--

At least one LLM provider key is required.

---

API Endpoints

All /v1/* endpoints require an API key via the X-API-Key header (unless noted otherwise).

Method	Path	Description	Auth
GET	/health	Health check -- returns status, version, active providers, uptime	No
GET	/v1/types	List supported document types and available providers	No
POST	/v1/parse	Parse a single document (file upload, text, or base64)	Yes
POST	/v1/parse/batch	Batch parse up to 20 files	Yes
POST	/v1/chat	Conversational AI (multi-LLM gateway)	Yes
POST	/v1/keys	Create a new API key (admin only)	Admin
GET	/v1/keys	List all API keys (admin only)	Admin
DELETE	/v1/keys/:name	Deactivate an API key by name (admin only)	Admin
GET	/v1/usage	Retrieve usage statistics (admin only)	Admin

---

Usage Examples

Parse a document (file upload)

cURL:

curl -X POST http://localhost:3000/v1/parse \
  -H "X-API-Key: YOUR_API_KEY" \
  -F "file=@receipt.jpg" \
  -F "type=receipt" \
  -F "quality=fast"

JavaScript:

const form = new FormData();
form.append('file', fileInput.files[0]);
form.append('type', 'receipt');
form.append('quality', 'fast');

const response = await fetch('http://localhost:3000/v1/parse', {
  method: 'POST',
  headers: { 'X-API-Key': 'YOUR_API_KEY' },
  body: form,
});

const result = await response.json();
console.log(result.data); // Structured JSON output

Parse a document (raw text)

cURL:

curl -X POST http://localhost:3000/v1/parse \
  -H "X-API-Key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "type": "invoice",
    "text": "Invoice #1042\nDate: 2025-03-15\nItem: Consulting — 10 hrs @ $150\nTotal: $1,500.00",
    "provider": "anthropic",
    "quality": "best"
  }'

JavaScript:

const response = await fetch('http://localhost:3000/v1/parse', {
  method: 'POST',
  headers: {
    'X-API-Key': 'YOUR_API_KEY',
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    type: 'invoice',
    text: 'Invoice #1042\nDate: 2025-03-15\nItem: Consulting — 10 hrs @ $150\nTotal: $1,500.00',
    provider: 'anthropic',
    quality: 'best',
  }),
});

const result = await response.json();
console.log(result.data);

Parse a document (base64)

curl -X POST http://localhost:3000/v1/parse \
  -H "X-API-Key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "type": "bank_statement",
    "image_base64": "'"$(base64 -w0 statement.png)"'",
    "image_mime": "image/png"
  }'

---

Document Types

Type	Key	What It Extracts
Receipt	receipt	Merchant, items, totals, VAT, payment method, category
Invoice	invoice	Line items, vendor/client details, payment terms, VAT
Bank Statement	bank_statement	Transactions with categories, balances, period summary
Contract	contract	Key terms, obligations, termination clauses, risk flags
General	general	Entities, key-value pairs, tables, summary (any document)

Set the type field in your request body. If omitted, defaults to general.

---

LLM Providers

Each provider offers two quality tiers (fast and best):

Provider	Fast Model	Best Model
Anthropic	Claude Haiku 4	Claude Sonnet 4
OpenAI	GPT-4o Mini	GPT-4o
Google	Gemini 2.0 Flash	Gemini 2.5 Flash

Specify "provider" and "quality" in the request body to control routing. If the requested provider is unavailable, the API falls back to the first configured provider.

---

Pricing Tiers

Tier	Monthly Requests	Rate Limit	Max File Size
Free	100 /day	10 /min	5 MB
Pro	5,000 /day	60 /min	25 MB
Enterprise	50,000 /day	300 /min	50 MB

API keys are created through the admin endpoint (POST /v1/keys) with the desired tier.

---

Deployment

Railway

A railway.json and Procfile are included for one-click deployment:

# Install the Railway CLI, then:
railway login
railway init
railway up

Set your environment variables in the Railway dashboard.

Docker

docker build -t docmind-api .
docker run -p 3000:3000 --env-file .env docmind-api

RapidAPI

DocMind API is available as a managed listing on RapidAPI. See RAPIDAPI_LISTING.md for integration details.

---

Project Structure

docmind-api/
  server.js               # Main application
  package.json             # Dependencies and scripts
  .env.example             # Environment variable template
  Dockerfile               # Container build
  Procfile                 # Railway / Heroku process file
  railway.json             # Railway deployment config
  rapidapi-middleware.js   # RapidAPI proxy-secret auth layer
  openapi.yaml             # OpenAPI 3.0 specification
  public/                  # Landing page assets
  index.html               # Documentation site (GitHub Pages)

---

Related Projects

• AI Native Solutions Hub -- central directory for all AI-native tools
• FallMesh -- distributed mesh networking
• FallGo -- Go-based backend toolkit
• sjgant80-hub on GitHub -- all repositories

---

License

MIT

---

Powered by Konomi Architecture · φ·κ=1
Member of the AI Craftspeople Guild