Convert Office files into LLM-ready packages.
Turn PPTX and DOCX files into local LLM-ready packages. XLSX support is planned.
Start with Office files
PPTX and DOCX converters are available now. The workspace generates a local LLM-ready ZIP package.
Files stay in the browser. The worker generates a ZIP package locally.
Upload a PPTX or DOCX file, then run the matched converter.
Preview panel
Inspect Markdown, HTML, or JSON before downloading.
Convert a supported Office file to inspect Markdown, HTML, and JSON output. After conversion, view full details in the Details workspace.
Supported outputs
Best for copying into ChatGPT, Claude, or Gemini.
Semantic sections with headings and image references.
Structured data for automation, agents, and RAG.
LLM entry index with slide summary, files, assets, and warnings.
Extracted images packaged under assets/.
Details workspace
Opened after conversion for structure summaries, previews, files, and warnings.
Convert a supported Office file to unlock details
After conversion, View details expands structure summaries, Markdown/HTML/JSON preview, package files, and warnings.
How it works
Four steps from Office upload to an LLM-ready ZIP package.
Upload Office file
Select a PowerPoint or Word file. The browser reads it locally.
Extract structure
Slides, document blocks, tables, and warnings become package data.
Preview outputs
Inspect Markdown, HTML, and JSON before downloading.
Download ZIP
Get Markdown, HTML, JSON, manifest, and assets together.
Privacy by default
Ready Office conversion runs locally in your browser. Your files do not need to be uploaded for the current workflow.
Files are processed by browser-side code for the ready Office workflow.
The MVP does not require accounts, cloud storage, or server-side task history.
Any server-enhanced workflow should be clearly labeled before upload.
FAQ
Does the file leave my browser?
The ready Office workflow runs locally and does not require a server upload.
What does the ZIP include?
Markdown, compact Markdown, HTML, JSON, manifest metadata, warnings, and extracted assets.
Are Markdown, HTML, and JSON pages different packages?
No. They produce the same ZIP package and only change the default preview tab.
Is DOCX or XLSX supported?
DOCX is ready for structured text extraction. XLSX is still planned.
Which output should I use with an LLM?
Use Markdown for direct reading, HTML for human review, and JSON for agents or RAG pipelines.