Dicklesworthstone/llm_aided_ocr

Health v2 · maintenance only

Flags: no_releases no_license

Adoption not part of the score

3001 stars · 215 forks observed · 2026-10-12

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-29, confidence not recorded

A Python tool that converts scanned PDFs to text via Tesseract OCR, then uses LLMs (local or API-based like OpenAI/Anthropic) to correct OCR errors, remove duplicates, and format output as clean Markdown. It includes smart chunking, image preprocessing, quality assessment, and optional GPU-accelerated local inference.

Use cases

• convert scanned pdfs to clean markdown
• fix OCR errors in digitized documents with an LLM
• correct tesseract output using local llama models
• turn old scanned letters or books into readable text
• remove headers and page numbers from OCR output
• batch process scanned pdfs into accurate text files

When to choose

• you have scanned PDFs whose raw OCR output is error-ridden and needs cleanup
• you want Markdown-formatted output from scanned documents
• you prefer running correction with a local LLM for privacy or cost reasons

When to avoid

• you need high-throughput OCR without LLM inference costs or latency
• you only need basic text extraction from clean digital PDFs
• you require a license-guaranteed dependency (license is non-standard)

Facets

cli-tool · maturity active

ocr nlp llm-inference pdf markdown image-processing pdf developer-tools large-language-models python cli tesseract ocr-correction pdf-to-markdown llm-post-processing document-digitization natural-language-processing linux macos gpu

Member repositories

For agents

markdown · JSON · MCP: product_card(name="Dicklesworthstone/llm_aided_ocr")