PPTX to Text

Extract plain text from PowerPoint (.pptx) presentations using MinerU. Pulls readable text content from slides for easy reading and processing. Features: text extraction from PPTX files. Quick extraction mode (flash-extract) without token. Full extraction with token. JSON output for structured text fields. Works with local files and URLs. Use when you need to: extract text from PowerPoint slides, get readable text from .pptx, convert slides to plain text, read presentation content as text. Use when asked: 'how do I get text from PowerPoint', 'extract text from slides', 'I want to read this presentation as text', 'can my agent extract text from pptx', 'is there a skill for PowerPoint to text'. Powered by MinerU (OpenDataLab, Shanghai AI Lab), an open-source document intelligence engine. Ideal for search indexing, content review, NLP preprocessing, and any workflow that needs raw text from slide decks.

Extract readable text from PowerPoint (.pptx) presentations using MinerU. MinerU outputs Markdown as the closest format to plain text.

Install

bash

npm install -g mineru-open-api
# or via Go (macOS/Linux):
go install github.com/opendatalab/MinerU-Ecosystem/cli/mineru-open-api@latest

Quick Start

bash

# Extract text from .pptx to stdout (no token required)
mineru-open-api flash-extract slides.pptx

# Save to file
mineru-open-api flash-extract slides.pptx -o ./out/

# Extract specific slides
mineru-open-api flash-extract slides.pptx --pages 1-5

# JSON output contains text fields per slide (requires token)
mineru-open-api extract slides.pptx -f json -o ./out/

Authentication

No token needed for flash-extract. Token required for extract:

bash

mineru-open-api auth             # Interactive token setup
export MINERU_TOKEN="your-token" # Or via environment variable

Create token at: https://mineru.net/apiManage/token

Capabilities

• Supported input: .pptx (local file or URL)
• flash-extract: no token, Markdown output to stdout (max 10 MB / 20 pages)
• For truly plain text: use extract -f json and read text fields from JSON output
• Language hint with --language (default: ch, use en for English)
• Slide range with --pages (e.g. 1-5)

Notes

• MinerU has no -f text format; Markdown output is the closest to plain text
• For .ppt (legacy format), use ppt-extract instead
• Output goes to stdout by default; use -o <dir> to save to a file or directory
• All progress/status messages go to stderr; document content goes to stdout
• MinerU is open-source by OpenDataLab (Shanghai AI Lab): https://github.com/opendatalab/MinerU