all2md

name	all2md
description	Convert, read, generate, search, and compare documents with the all2md CLI and Python library. Use whenever a task involves reading or extracting text/tables from a document (PDF, Word, PowerPoint, Excel, HTML, email, EPUB, Jupyter, images, or 100+ source-code and text file types); converting between formats (PDF→Markdown, Markdown→DOCX/PDF/PPTX/HTML/EPUB, any-to-any); generating documents or static sites from Markdown; grepping for patterns inside non-plaintext files; ranked keyword or semantic search across a document collection; or diffing two documents. Triggers include 'read this PDF', 'extract text/tables', 'convert to Word/PDF', 'make slides', 'export to EPUB', 'search these documents', 'grep this docx', or 'what changed between these files'.
metadata	{"author":"all2md","version":"2.0"}

all2md

all2md converts between 40+ document formats and Markdown, from both the command line and a Python API. It reads PDF, DOCX, PPTX, XLSX, HTML, EML, EPUB, ODT/ODP/ODS, RTF, Jupyter notebooks, images, and 100+ source-code and config file types; renders Markdown back out to DOCX, PDF, PPTX, HTML, EPUB, LaTeX, RST, and more; and ships task commands for search, grep, diff, and static-site generation.

This file is the overview and index. Pick the task below and read the matching reference file for full CLI flags, Python API, and examples.

Choose a reference

You want to…	Read	Command family
Read / extract text or tables from a document	references/read.md	all2md <file>
Convert a file from one format to another	references/convert.md	all2md <file> --output-format <fmt>
Generate a document or static site from Markdown	references/generate.md	all2md <md> --output-format <fmt>, generate-site, arxiv
Find a pattern inside documents (like grep)	references/grep.md	all2md grep
Ranked keyword/semantic search across a collection	references/search.md	all2md search
Compare two documents / see what changed	references/diff.md	all2md diff
Chunk a document for RAG/LLM pipelines (JSONL + provenance)	references/chunk.md	all2md chunk

Core model

• Read: to_markdown(path) / to_ast(path) auto-detect the input format and route to the right parser.
• Write: from_markdown(...) / from_ast(doc, fmt) render an AST to a target format.
• CLI: all2md <input> [--output-format FMT] [-o OUT]. Default output is Markdown to stdout. Use - for stdin.
• Format-specific flags follow the pattern --<format>-<option> for parsers (e.g. --pdf-pages) and --<format>-renderer-<option> for renderers (e.g. --html-renderer-no-standalone).

Quick start

# Read any document to Markdown (stdout)
all2md report.pdf

# Save to a file
all2md report.pdf -o report.md

# Convert between formats
all2md report.md --output-format docx -o report.docx
all2md page.html --output-format pdf -o page.pdf

# Read from stdin
cat report.pdf | all2md -

# Search, grep, diff
all2md grep "revenue" *.pdf
all2md search "machine learning" ./papers --vector
all2md diff v1.docx v2.docx

from all2md import to_markdown, from_markdown, to_ast, from_ast

markdown = to_markdown("report.pdf")          # read
from_markdown("report.md", "docx", output="report.docx")   # generate
doc = to_ast("report.pdf")                    # AST for programmatic access
from_ast(doc, "html", output="report.html")   # render

Tips

• all2md list-formats shows every input/output format with dependency status.
• all2md check-deps verifies optional dependencies (PyMuPDF for PDF, etc.).
• Beyond the task families above, the CLI also has: all2md view <file> (preview in a browser), all2md serve (local HTTP API), all2md edit <file> (web editor), all2md lint <file> (Markdown linter with --fix), and all2md config generate|show (configuration files). Run all2md <command> --help for flags.
• all2md --help full prints the complete flag reference, including all format-specific options.
• all2md llm-help [topic] prints these references to stdout (topics: read, convert, generate, grep, search, diff, chunk) — handy when driving the CLI without installing the skill.
• Every CLI option also reads an environment variable: ALL2MD_<OPTION> (e.g. ALL2MD_PDF_OCR_ENABLED=true).