Skip to content
Stork.AIstork.ai
AI ToolSleeping Giant

The Vesuvius Challenge Review

The Vesuvius Challenge is a global competition that crowdsources the development of AI algorithms to decipher ancient Herculaneum scrolls.

shipped Jul 31, 2026aifreemium
ai
The Vesuvius Challenge — product screenshot

Why it matters

1Over $2 million in open prizes, with $20,000 awarded monthly.
2Successfully read PHerc. 1667 (Scroll 4) in full by June 2026, recovering 20 columns of text.
3Optimized X-ray scanning protocols in 2025, reducing scan time for large scrolls to under two hours.
4Approximately 70% of PHerc. 172 (Scroll 5) digitally unwrapped by January 2026.

Stork Quadrant

Sleeping Giant· 42/100

Has a real moat but invisible to agents. Add an MCP and you'd climb.

“This is not a software tool — it's a coordinated scientific prize competition built around physically unique artifacts that no LLM can replicate. The CT scan data of the actual scrolls is proprietary and irreplaceable. The crowdsourced network of researchers, the prize structure, and the brand authority in the digital humanities world compound over time. An LLM alone cannot unseal a carbonized scroll.”

— Claude Sonnet 4.6, scored 2026-07-31

Defensibility · 73/100

  • Physical-world coupling
  • Regulatory moat
  • Network liquidity
  • Proprietary refreshing data
  • High-trust catastrophic workflows
  • Multi-party coordination
  • Brand / community / taste

An LLM alone could replace

  • Generate hypotheses about ancient Latin or Greek text content given partial transcriptions
  • Summarize known scholarship on Herculaneum scrolls and Epicurean philosophy
  • Write image segmentation or ML pipeline code for papyrus analysis
  • Explain historical context of the Villa of the Papyri

Agent-Readiness · 5/100

  • Verified MCP
  • Listed on agent surfaces
  • Usage-based pricing
  • Headless agent auth
  • Public OpenAPI
  • Active changelog
  • llms.txt— https://scrollprize.org/llms.txt

How to defend

Keep the raw scan data and ground-truth annotations locked inside the competition infrastructure so the dataset itself becomes the moat — then license access to institutions and build the coordination layer between papyrologists, ML researchers, and museums.

  • Ship an MCP server and list it on Stork — biggest single point gain (+25).
  • Get listed in the Anthropic MCP registry, Cursor, or Claude Desktop (+20).
  • Add a usage-based or per-call tier; per-seat-only pricing dies when agents replace seats (+15).
  • Expose API-key auth with a self-serve sandbox tier; remove sales-call gates (+15).
  • Publish an OpenAPI spec at /openapi.json or /.well-known/openapi (+10).
How this score is computed →See the full quadrantHow to defend

About The Vesuvius Challenge

Founded
2023
Team Size
20+
Funding
Bootstrapped
Target Audience
Researchers, machine learning practitioners, and enthusiasts of historical texts.

Leadership

Nat FriedmanInstigator, Director & Founding SponsorLinkedIn
Daniel GrossFounding SponsorLinkedIn
Brent SealesPrincipal Advisor, PhDLinkedIn
GitHub

Specs

API Docs

View Documentation →

GitHub

View Repository →

API Available

Yes, public API

overview

What is The Vesuvius Challenge?

The Vesuvius Challenge is a machine learning, computer vision, and geometry tool developed by Nat Friedman, Daniel Gross, and Brent Seales that enables researchers, machine learning practitioners, and enthusiasts of historical texts to virtually unroll and read ancient carbonized Herculaneum scrolls. It employs micro-CT scanning and AI models to detect ink traces within fragile, physically unopenable papyri, recovering lost ancient Greek and Latin texts.

The challenge focuses on deciphering the Herculaneum scrolls, which were carbonized during the 79 AD eruption of Mount Vesuvius. These scrolls are too delicate for traditional physical unwrapping methods. The project utilizes high-energy synchrotron X-ray scans to create detailed 3D images, with AI algorithms trained to identify subtle ink marks. Open-source software like VC3D facilitates interactive segmentation and virtual unwrapping, contributing to the recovery of ancient philosophy, literature, and history.

features

Key Features of The Vesuvius Challenge

The Vesuvius Challenge incorporates several key features designed to facilitate the virtual unwrapping and reading of ancient Herculaneum scrolls through a collaborative, AI-driven approach.

  • Machine learning competition for ink detection and text decipherment.
  • Computer vision application for 3D volume reconstruction and papyrus segmentation.
  • Open prizes totaling over $2,140,000 for breakthroughs in scroll reading.
  • Focus on non-destructive virtual unwrapping of carbonized Herculaneum scrolls.
  • Collaboration with academic institutions and leading X-ray facilities like ESRF.
  • Crowdsources AI algorithm development from a global community.
  • Publicly available scans, data, code, and internal models for open science.
  • Utilizes advanced X-ray scanning techniques (tetra-, hexa-, and octa-helical acquisitions).
  • Supports virtual unrolling and reading of ancient carbonized scrolls.

use cases

Who Should Use The Vesuvius Challenge?

The Vesuvius Challenge targets a diverse group of individuals and organizations interested in advancing AI, historical research, and open science. Its structure as a competition and open-source project makes it accessible to various participants.

  • Machine learning engineers: For developing and refining models for ink detection and text segmentation.
  • Computer vision specialists: For creating tools for 3D volume reconstruction and papyrus surface identification.
  • Geometers and software developers: For contributing to virtual unwrapping algorithms and interactive software like VC3D.
  • Papyrologists and classical scholars: For analyzing recovered texts and contributing to historical discovery.
  • Students and researchers: For participating in a high-impact open science project with monetary incentives.

how to use

How to Use The Vesuvius Challenge

Participating in The Vesuvius Challenge involves engaging with its open-source data, developing algorithms, and submitting solutions to ongoing prize categories. The project provides resources and a community for collaboration.

  • 1Access the official website (scrollprize.org) to review current prize categories and rules.
  • 2Download publicly available micro-CT scans and associated data of the Herculaneum scrolls.
  • 3Utilize open-source tools like VC3D for interactive segmentation and virtual unwrapping.
  • 4Develop machine learning models for ink detection or computer vision algorithms for surface reconstruction.
  • 5Submit developed algorithms or identified text segments according to specific prize guidelines.
  • 6Engage with the community on platforms like Discord for collaboration and support.

pricing

The Vesuvius Challenge Pricing & Plans

The Vesuvius Challenge operates on a freemium model, providing free access to core data and tools while offering significant monetary prizes for successful contributions. There are no direct subscription costs for participation.

  • Freemium: Free access to all scans, data, code, and internal models for development and research.
  • Prizes: Over $2 million in open prizes, including a $1 million Grand Prize for 2027 and $20,000 awarded monthly for ink detection and text recovery.

Pros

  • +Enables non-destructive recovery of ancient texts from fragile Herculaneum scrolls.
  • +Fosters a highly collaborative global community for open science and innovation.
  • +Offers substantial monetary prizes (over $2 million) to incentivize breakthroughs.
  • +Provides public access to all scans, data, code, and internal models.
  • +Significantly advances machine learning, computer vision, and geometry fields.
  • +Accelerates historical discovery by recovering previously unreadable ancient texts.

Cons

  • −Requires specialized knowledge in machine learning, computer vision, or geometry to contribute effectively.
  • −Success is dependent on the continued development and refinement of AI algorithms.
  • −The process of fully reading all 300 extant scrolls remains a long-term, resource-intensive endeavor.
  • −The project's funding model relies on sponsorships and prize money, which may fluctuate.
Visit The Vesuvius Challenge↗
Connect
𝕏
X / Twitterx.com/scrollprize
⌘
GitHubgithub.com/ScrollPrize
in
LinkedInwww.linkedin.com/in/dr-hendrik-schilling-a2019418a
💬
Discorddiscord.gg/V4fJhvtaQn

Featured in articles