The MCP Mistral OCR server provides optical character recognition capabilities by leveraging Mistral AI's OCR API. It processes both local files and URLs, supporting various image formats (JPG, PNG, WebP) and PDFs, with results saved as timestamped JSON files. The implementation includes Docker containerization for easy deployment, configurable environment variables for API authentication and file storage, and offers two primary tools: process_local_file for handling files from a configured directory and process_url_file for processing remote documents with explicit file type specification. This server is particularly useful for workflows requiring text extraction from visual documents within AI assistant conversations.
暂无评论. 成为第一个评论的人!
登录以参与讨论
Process a file from the configured OCR_DIR directory. Parameters: filename (string)
Process a file from a URL. Requires explicit file type specification. Parameters: url (string), file_type (string - either 'image' or 'pdf')