Skip to content

index

Jump to bottom
wiki[bot] edited this page Apr 3, 2026 · 6 revisions

@CKIR.IO/VISIONS

@CKIR.IO/VISIONS allows uploading one or more images (PNG, JPG, JPEG, WEBP) and supports AI-driven description, comparison and text extraction via OCR. It exposes multiple REST endpoints and implements the Model Context Protocol (MCP) using the JSON‑RPC 2.0 transport.

Features

  • Image Description - AI-powered image analysis and description
  • Image Comparison - Compare multiple images and identify differences
  • OCR - Extract text from images using vision models
  • REST API - Simple multipart upload endpoint
  • MCP Protocol - Model Context Protocol support with JSON-RPC 2.0
  • Real-time Results - Socket.IO for streaming responses

Installation

# Install dependencies
pnpm install

# Start in development mode
pnpm start:dev

Documentation

  • REST API - Multipart endpoint with task types: describe, compare, ocr
  • MCP - JSON-RPC 2.0 endpoint with tools/list and visions.analyze tool
  • Socket.IO - Real-time streaming of analysis results
  • Docker Compose - Docker setup
  • System Environment - Environment variables

Clone this wiki locally