dokyumi

AI-powered document parsing from the command line. Extract structured data from PDFs, images, tax forms, invoices, receipts, and more.

npm install -g dokyumi

Quick Start

# Set your API key
dokyumi login dk_live_abc123

# Parse a document
dokyumi parse invoice.pdf

# Parse with a specific schema
dokyumi parse 1099.pdf --schema 1099-nec

# Batch parse a folder
dokyumi batch ./tax-docs/ --output results.json

# Pipe to jq
dokyumi parse w2.pdf --quiet | jq '.data.wages'

What It Does

Upload any document — PDF, image, scan, photo — and get structured JSON data back in seconds. Powered by AI that reads documents the way humans do.

Supported documents: 1099s, W-2s, 1040s, invoices, receipts, bank statements, medical bills, EOBs, leases, insurance dec pages, contracts, and more.

Supported formats: PDF, PNG, JPG, TIFF, WebP, GIF

Commands

| Command | Description | |---------|-------------| | dokyumi parse <file> | Parse a single document | | dokyumi batch <dir> | Parse all documents in a directory | | dokyumi schemas | List your available schemas | | dokyumi login <key> | Save your API key | | dokyumi whoami | Check authentication status |

Options

| Flag | Description | |------|-------------| | --schema <name> | Use a specific parsing schema | | --output <file> | Write output to file | | --format <json\|csv\|table> | Output format (default: json) | | --quiet | Suppress progress output |

Authentication

Get your API key at dokyumi.com/dashboard/keys.

# Option 1: Login command (saves to ~/.dokyumi/config.json)
dokyumi login dk_live_your_api_key

# Option 2: Environment variable
export DOKYUMI_API_KEY=dk_live_your_api_key

Use with AI Agents

Dokyumi is designed to work seamlessly with AI agents and automation:

# In a shell script
RESULT=$(dokyumi parse receipt.pdf --quiet)
TOTAL=$(echo $RESULT | jq -r '.data.total')

# In a Makefile
parse:
  dokyumi batch ./incoming/ --output parsed.json

# With Claude/ChatGPT via MCP
# See: dokyumi-mcp package

Pricing

• Free: 50 extractions/month
• Starter ($29/mo): 500 extractions
• Growth ($99/mo): 2,000 extractions
• Enterprise ($399/mo): 10,000 extractions

View pricing →

Links

• Documentation
• Dashboard
• MCP Server — Use with Claude, ChatGPT, and other LLMs
• OpenClaw Skill — Use with OpenClaw agents

License

MIT