Process Document with Claude OCR — claude_ocr_process

This function processes a local PDF or image file using Claude's document understanding capabilities and returns structured OCR results.

Usage

claude_ocr_process_file(
  file_path,
  api_key = Sys.getenv("ANTHROPIC_API_KEY"),
  model = "claude-opus-4-6",
  max_tokens = 16000,
  extraction_prompt = NULL,
  endpoint = "https://api.anthropic.com/v1/messages"
)

Arguments

file_path: Character string. Path to a local PDF, PNG, JPEG, or other supported image file.
api_key: Character string. The Anthropic API key. Default is to retrieve from environment variable "ANTHROPIC_API_KEY".
model: Character string. The Claude model to use. Default is "claude-opus-4.5-20250514".
max_tokens: Integer. Maximum tokens in response. Default is 16000.
extraction_prompt: Character string. Custom prompt for extraction. If NULL, uses default Tensorlake-compatible prompt.
endpoint: Character string. The Claude Messages API endpoint. Default is "https://api.anthropic.com/v1/messages".

Value

A list containing the Claude API response with structured OCR data.

Author

Nathan C. Layman