Gemina Filetag
Description
# Gemina FileTag Tag, rename, and enrich PDFs and images with structured metadata extracted via OCR + LLM. Built for AI agents that need to make sense of unstructured document piles. ## What it does - **Tag** — extract structured fields (title, date, party names, document type…
About
# Gemina FileTag Tag, rename, and enrich PDFs and images with structured metadata extracted via OCR + LLM. Built for AI agents that need to make sense of unstructured document piles. ## What it does - **Tag** — extract structured fields (title, date, party names, document type, language, summary, etc.) from any PDF or…
Details
- Author
- tommyil
- Downloads
- 301
- Categories
- Other, File Management, AI
Jump to
- Extract structured fields (title, date, party names, document type, language, summary) from PDFs or images.
- Generate descriptive filenames from extracted metadata.
- Embed metadata back into the file (PDF info dictionary, EXIF).
- Zero setup with a free tier of 1,500 tags per month, no credit card required.
- Document-type agnostic — works with invoices, receipts, contracts, IDs, screenshots, handwritten notes.
- Returns a predictable schema every time, with null values when unsure.
Setting up with Highlight
This MCP is not yet compatible with Highlight’s one-click setup. However, you can still use it with Highlight by following these steps:
- Download and install Highlight from highlightai.com/download
- Navigate to the plugins tab and select "Add Custom Plugin"
-
Configure the plugin with the settings below
Plugin Name
Gemina FiletagCommand (node, npx, python, etc.)Please refer to the README for specific instructions on how to obtain API keys or other required environment variables.
- Enable "Start Automatically" if you want the plugin to start when Highlight launches
From the repository
Configure Gemina Filetag as a Streamable HTTP MCP server by adding the URL https://api.gemina.co/api/v1/mcp/ and your API key (obtained free from console.gemina.co) to the mcpServers section of your client. No local installation, Docker, or model downloads are needed.
Claude Desktop / Cursor
Paste into your MCP client config file to install this server.
{
"mcpServers": {
"gemina filetag": {
"gemina": {
"type": "streamableHttp",
"url": "https://api.gemina.co/api/v1/mcp/",
"headers": {
"X-API-Key": "YOUR_GEMINA_API_KEY"
}
}
}
}
}McpServers
{
"gemina": {
"type": "streamableHttp",
"url": "https://api.gemina.co/api/v1/mcp/",
"headers": {
"X-API-Key": "YOUR_GEMINA_API_KEY"
}
}
}Gemina FileTag
Tag, rename, and enrich PDFs and images with structured metadata extracted via OCR + LLM. Built for AI agents that need to make sense of unstructured document piles. ## What it does - Tag — extract structured fields (title, date, party names, document type, language, summary, etc.) from any PDF or image. - Rename — generate descriptive filenames from extracted metadata. - Enrich — embed the metadata back into the file itself (PDF info dictionary, EXIF) so it stays with the file forever. ## Why use it - Zero setup, free tier — 1,500 tags/month, no credit card. - Hosted Streamable HTTP — no local install, no Docker, no model downloads. - Document-type agnostic — invoices, receipts, contracts, IDs, screenshots, handwritten notes. - Predictable schema — returns the same field shape every time; null when unsure. ## Quick start ```json { "mcpServers": { "gemina": { "type": "streamableHttp", "url": "https://api.gemina.co/api/v1/mcp/", "headers": { "X-API-Key": "YOUR_GEMINA_API_KEY" } } } } Get a free API key at console.gemina.co (https://console.gemina.co/registration/create-account) — takes under a minute, no credit card. Pricing ┌──────┬────────────┬─────────────┐ │ Tier │ Tags/month │ Cost │ ├──────┼────────────┼─────────────┤ │ Free │ 1,500 │ $0 │ ├──────┼────────────┼─────────────┤ │ Paid │ unlimited │ usage-based │ └──────┴────────────┴─────────────┘ Links - Website: https://www.gemina.co/filetag - GitHub: https://github.com/tommyil/gemina-mcp - Console: https://console.gemina.co - MCP Registry: co.gemina/filetagSign in to leave a review
Use Google, GitHub, or an email account so ratings stay tied to real people.
No reviews posted yet.
