open-index/open-markdown-v2

Open Markdown

Clean markdown from the web, ready for training and retrieval

What is it?

Open Markdown is a large-scale web text dataset built from Common Crawl. Common Crawl is a non-profit that crawls the web and freely provides its archives to the public. Every page goes through a pipeline that extracts the main content from raw HTML, converts it to clean Markdown, and packages the result into Parquet files with WARC metadata for traceability.

The dataset currently includes crawl CC-MAIN-2026-25 with ~2,028,365,882 documents across 100,000 shards. Processed ~331.0 TB of raw HTML into ~9.5 TB of clean Markdown, a 97.1% reduction. We plan to add more snapshots over time.

Open Markdown is released under the Open Data Commons Attribution License (ODC-By) v1.0, the same license used by Common Crawl.

What is being released?

Each Common Crawl WARC file (about 1 GB of compressed HTML) becomes one Parquet shard. The shards live under a crawl-specific, Hive-partitioned directory so multiple snapshots can coexist and partition-aware tools can filter by crawl without a full scan:

data/
  crawl=CC-MAIN-2026-25/
    000000.parquet
    000001.parquet
    000002.parquet
    ...

Every row in a Parquet file is one web page. Each row keeps the warc_record_id parsed from the original WARC header so you can trace a document back to its source record, plus html_length and markdown_length to measure the compression from raw HTML to clean Markdown.

How to download and use Open Markdown

Using datasets

from datasets import load_dataset

# stream the entire dataset
ds = load_dataset("open-index/open-markdown-v2", name="CC-MAIN-2026-25", split="train", streaming=True)
for doc in ds:
    print(doc["url"], doc["markdown_length"])

# load a single shard into memory
ds = load_dataset(
    "open-index/open-markdown-v2",
    data_files="data/crawl=CC-MAIN-2026-25/000000.parquet",
    split="train",
)

Using huggingface_hub

from huggingface_hub import snapshot_download

folder = snapshot_download(
    "open-index/open-markdown-v2",
    repo_type="dataset",
    local_dir="./open-markdown-v2/",
    allow_patterns="data/crawl=CC-MAIN-2026-25/**/*.parquet",
)

For faster downloads, install pip install huggingface_hub[hf_transfer] and set HF_HUB_ENABLE_HF_TRANSFER=1.

Using DuckDB

SELECT url, host, markdown_length
FROM read_parquet('hf://datasets/open-index/open-markdown-v2/data/crawl=CC-MAIN-2026-25/**/*.parquet')
WHERE host = 'en.wikipedia.org'
LIMIT 10;

Dataset card for Open Markdown

Dataset Structure

Data Instance

The following is an example row from the dataset:

{
  "doc_id": "6aaa5be7a9175105aa60e39ea1d087fc",
  "url": "https://example.com/article/interesting-topic",
  "host": "example.com",
  "crawl_date": "2026-06-12",
  "warc_record_id": "<urn:uuid:a1b2c3d4-e5f6-7890-abcd-ef1234567890>",
  "html_length": 48210,
  "markdown_length": 3847,
  "markdown": "# Interesting Topic\n\nThis is the main content of the page..."
}

Data Fields

Column	Type	Description
doc_id	string	Deterministic SHA-256 hash of the URL, first 16 bytes in hex. Identical URLs always produce the same doc_id across crawls, so cross-crawl dedup is an equi-join on this column
url	string	Original URL of the crawled page
host	string	Hostname extracted from the URL
crawl_date	string	Date of the WARC record (YYYY-MM-DD)
warc_record_id	string	WARC-Record-ID of the original HTTP response (<urn:uuid:...>)
html_length	int64	Byte length of the original HTML body before conversion
markdown_length	int64	Byte length of the converted Markdown body
markdown	string	Clean Markdown content extracted from the page

Data Splits

The default subset includes all available data across all crawl snapshots. You can also load a specific crawl by using its ID as the config name (for example CC-MAIN-2026-25).

Dataset Creation

Curation Rationale

Most open web datasets either release raw text without structure or keep the HTML and leave parsing to the user. Open Markdown sits in between: it converts every page to Markdown so the content is immediately usable for training and retrieval, while preserving the warc_record_id so you can always trace back to the source record.

Source Data

The source data consists of web pages crawled by the Common Crawl foundation. Common Crawl archives billions of pages across the public web and makes the raw WARC files freely available on Amazon S3.

Data Processing Steps

The processing pipeline runs as a single streaming pass with no intermediate files:

1. Download raw .warc.gz files from Common Crawl S3 (each file is roughly 1 GB compressed)
2. Filter to keep only HTTP 200 responses with a text/html content type, discarding images, scripts, redirects, and error pages
3. Convert HTML to clean Markdown. go-trafilatura tuned for recall (FavorRecall, with Readability and DomDistiller fallbacks) isolates the main content node and strips navigation, ads, and boilerplate, then a direct node-tree walk renders GitHub-flavored Markdown with links resolved to absolute URLs
4. Export straight to Apache Parquet with Zstd compression

The pipeline streams from the compressed WARC through conversion directly into Parquet. Pages that produce empty conversions are dropped.

Compression Ratios

Numbers below are measured across 170 shards of CC-MAIN-2026-25 (3,448,222 pages) and projected to the full crawl of 100,000 WARC files.

Stage	Measured	Projected (100,000 files)	Reduction
Raw WARC (.warc.gz, downloaded)	145.6 GB	~83.6 TB	—
HTML extracted (uncompressed)	576.3 GB	~331.0 TB	—
Markdown (clean text)	16.5 GB	~9.5 TB	-97.1% vs HTML
Final Parquet (Zstd)	8.4 GB	~4.8 TB	-49.2% vs Markdown

The big win is the HTML to Markdown conversion: trafilatura keeps only the main content and the renderer drops every tag, script, style, and navigation block. This cuts ~331.0 TB of uncompressed HTML down to ~9.5 TB of Markdown, a 97.1% reduction. Parquet with Zstd then compresses the Markdown further.

Processing Times

Pipeline timings across 170 shards of CC-MAIN-2026-25:

Convert  (stream WARC to Markdown to Parquet)    ████████████████████████  36h 58m 3s
Publish  (HuggingFace upload)                    █░░░░░░░░░░░░░░░░░░░░░░░  7m 53s

Personal and Sensitive Information

No additional PII filtering is applied beyond what Common Crawl provides. As the dataset is sourced from the public web, it is likely that some personally identifiable information is present. If you find your own PII in the dataset and would like it removed, please open an issue on the repository.

Considerations for Using the Data

Social Impact

By releasing both the dataset and the full processing pipeline, we aim to lower the barrier to training and evaluating language models on high quality web data. Researchers and practitioners who cannot afford to run their own Common Crawl processing pipelines can use Open Markdown directly.

Discussion of Biases

Open Markdown inherits the biases present in Common Crawl and the public web at large. The trafilatura extraction step favors article-like pages and may underrepresent content from forums, social media, and non-standard page layouts. We have not applied any machine-learning-based quality or toxicity filters, as such filters have been shown to disproportionately remove content from certain dialects and communities.

Known Limitations

Code-heavy pages may not convert well to Markdown. If you are training a model that needs strong code performance, consider supplementing Open Markdown with a dedicated code dataset such as The Stack v2. Similarly, highly structured pages like Wikipedia may have better formatting in dedicated Wikipedia dumps than in their Common Crawl versions.

Additional Information

Licensing

The dataset is released under the Open Data Commons Attribution License (ODC-By) v1.0. The use of this dataset is also subject to Common Crawl's Terms of Use. The original content remains subject to the rights and terms of its respective publishers.

Contact

Please open a discussion on the Community tab for questions, feedback, or issues.