News and updates from the Wikimedia Enterprise team
-
Working on Openness and Trust in AI Wikimedia Enterprise invites you to a special social event at NeurIPS 2025 in San Diego this December. This in-person event will explore the intersections between generative AI data and open, trusted datasets. Representatives from the Wikimedia Foundation (Enterprise team), MLCommons, and the AI Alliance will share insights into…
-
Access Wikipedia’s most valuable tables as structured JSON with the new Parsed Tables feature from Wikimedia Enterprise. Instantly convert complex tables into clean, machine-readable data without scraping. Enhance your AI, search, and knowledge graph projects with reliable, human-curated facts that were previously locked away in HTML and wikitext.
-
Nomic AI used Wikimedia Enterprise’s Structured Contents dataset, via Hugging Face, to build the first full open-source vectorization of multilingual Wikipedia. Their work highlights how structured open data can accelerate AI research, improve model performance, and enable new forms of data visualization.
-
Explore Wikipedia content in a clean, structured format with our new beta dataset on Kaggle. Built from our Snapshot API using the Structured Contents beta, it’s ideal for data science, ML training, and experimentation.
-
Wikimedia Enterprise is partnering with ProRata.ai to power its new search engine, Gist.ai, with reliable, human-curated Wikimedia content. The collaboration supports a sustainable content ecosystem through transparent attribution and API-driven innovation—ensuring creators are credited and content remains discoverable in the AI era.
-
The latest API release boosts Wikipedia data integration with parsed references in JSON and two quality scoring models – Reference Need and Reference Risk. These enhancements streamline citation access and improve content reliability for developers.
-
Wikimedia Enterprise joined Creative Commons at SXSW 2025 on March 9th in downtown Austin for a day of insightful conversations and panels exploring the intersection of artificial intelligence and open data. The discussions emphasized the critical importance of protecting and ethically advancing open access principles amidst rapid technological growth. The event brought together industry leaders,…
-
Wikimedia Enterprise and Pleias are partnering to drive ethical AI innovation with high-quality, structured data. By integrating Wikimedia’s verifiable datasets, Pleias enhances its AI models while ensuring openness, auditability, and multilingual accuracy.
-
A Partnership for Sustainability of Knowledge and the Planet – Wikimedia Enterprise has announced an exciting new partnership with Ecosia search engine.
-
Discover how Wikimedia Enterprise APIs transformed in 2024 with parsed article sections, an AI focus, and greater free access. Learn about our enhanced developer tools and how to leverage Wikipedia’s dynamic content with recurring API access and improved data structures.
-
We’re releasing an early beta dataset on Hugging Face, offering structured content from English and French Wikipedia. This machine-readable dataset, derived from our Snapshot API’s new Structured Contents beta, opens up new possibilities for AI and machine learning applications.
-
Snapshot API now includes a beta Structured Contents endpoint, offering bulk access to parsed Wikipedia data for testing partners.
-
We’re excited to announce a major upgrade to our free API tier. Users now benefit from recurring monthly credits, replacing the previous lifetime limit, and enjoy twice the speed in data updates. These enhancements are designed to provide more value and flexibility to our free account holders.
-
Unlock deeper insights into Wikipedia content with Wikimedia Enterprise’s Credibility Signals. Learn how these tools provide critical context, empowering informed decisions for AI models, knowledge graphs, and beyond.
-
Discover how to leverage the Wikimedia Enterprise On-demand API to improve the contextual accuracy of open-source language models like Meta’s Llama 3. Our tutorial walks you through setting up a local RAG-based application for more accurate AI responses.
-
In this engineering tutorial, we show a simple way to build a working knowledge panel pulling pre-parsed content from Wikipedia articles using Wikimedia Enterprise API Structured Contents endpoint.
-
Wikimedia Enterprise Realtime API has two new significant features: Parallel Connections and Restart support, to help streamline ingesting over one million daily events across all supported projects.
-
Our APIs have been updated with new metadata about every article edit along with a Probability of Revert helping you make more informed decisions. The Version object has some impressive updates we’re excited to show you.
-
We’ve expanded the data available in our On-Demand Structured Contents endpoint by introducing two significant features: Article Body Sections and short Descriptions.
-
We’ve heard all your requests for a more machine-readable API for Wikimedia data. We are announcing a new Structured Contents endpoint with the fully parsed contents of Wikipedia article Infoboxes in JSON! Jump into the article to read about it and get started.