Web data API that turns pages into Markdown or JSON for AI models, with search, crawl and natural-language page interaction.

Marker overview and key facts (October 2026)
Open-source tool from New York-based Datalab that converts PDF, Office and images into Markdown, JSON and HTML, optionally with an LLM.
from $400 per month
What is Marker?
Marker is an open-source converter that turns PDFs, Office files, images and e-books into Markdown, JSON, chunks and HTML for use with language models. Datalab, its maintainer, also sells a hosted API and on-premises licenses.
Marker is maintained by Datalab, based in Manhattan, New York. It converts PDF, image, PPTX, DOCX, XLSX, HTML and EPUB files in all languages, extracts images and runs on GPU, CPU or Apple MPS. The code is Apache 2.0; the model weights use a modified OpenRAIL-M license that is free for research, personal use and startups under $5 million in funding or revenue. As of September 28, 2026 the repository had about 40,000 GitHub stars, with a last push on September 13, 2026. Datalab's hosted platform runs its newer Chandra model and lists SOC 2 Type II.
Marker AI features and use cases
- Converts documents into clean MarkdownModels detect layout and reading order and format tables, forms, equations, links and code blocks. The output can go straight into a RAG index or a prompt.
- Cleans up page clutterHeaders, footers and other artifacts are removed during conversion. Search results and chunks contain the content instead of repeated page furniture.
- Improves accuracy with an LLMAn optional mode sends difficult parts to a language model, with a prompt you can adjust. Teams can raise quality on hard documents where it matters.
What sets Marker apart
- Separate licenses for code (Apache 2.0) and model weights (modified OpenRAIL-M with a revenue threshold).
- Hosted API with published per-page prices and on-premises, air-gapped options for Enterprise.
- Runs on GPU, CPU or Apple silicon.
Where Marker is strongest
Datalab is based in Manhattan, New York.
Questions to ask before you buy Marker
Check whether your company exceeds the $5 million funding or revenue threshold, because then commercial use of the model weights needs a paid license from Datalab. For the hosted API, a BAA and DPA are included from the Team plan, and SSO is an Enterprise feature. Self-hosting means your team maintains the models and hardware.
Marker reviews
No reviews yet. Be the first to share how your team uses it.
Every review shows its verification level. We read each one before it goes live. How we check reviews
Questions about Marker
Is Marker free for companies?
The code is Apache 2.0. The model weights are free for research, personal use and startups under $5 million in funding or revenue; larger companies need a commercial license.
What does the hosted API cost?
Datalab lists a free plan with a $20 monthly allowance for work email, a Team plan at $400 per month and per-page rates such as $4 per 1,000 pages for fast conversion.
Which files can Marker convert?
PDF, images, PPTX, DOCX, XLSX, HTML and EPUB, in all languages.
Marker update historyno updates yet
No changes recorded for Marker yet.
Changes the vendor made and confirmed user reports, each checked against the vendor's own pages. We review every tool monthly.
Alternatives to Marker
Tools in the same category (Document parsers) with the most overlapping markets.
Document parsing API from LlamaIndex that prepares more than 90 formats for AI with layout awareness, with free credits and an EU region.
AWS OCR and document analysis service that extracts text, handwriting, tables, forms and expense data from scans.
Paris-based OCR and extraction API that turns invoices, receipts, IDs, resumes and custom document types into structured JSON.



