Vesuvius Challenge
Vesuvius Challenge is a nonprofit, crowdsourced research initiative that uses machine learning, computer vision, and geometry processing to virtually unwrap and read carbonized Herculaneum papyrus scrolls sealed since 79 AD, serving volunteer ML researchers and classical papyrologists worldwide.
- Company typePrivate
- Founded2023
- Headquarters—
- Headcount11–50
- GTM typeB2B
- OfferingServices
Vesuvius Challenge firmographics
Firmographics- Name
- Vesuvius Challenge
- Legal name
- Vesuvius Challenge
- Website
- https://scrollprize.org
- Company type
- Private
- Founded year
- 2023
- Operating status
- Operating
- Headcount range
- 11–50 employees
- Short description
- Vesuvius Challenge is a nonprofit, crowdsourced research initiative that uses machine learning, computer vision, and geometry processing to virtually unwrap and read carbonized Herculaneum papyrus scrolls sealed since 79 AD, serving volunteer ML researchers and classical papyrologists worldwide.
- Ownership category
- akta.pro rank
Vesuvius Challenge industry classification
Industry- Product category
- Digital Humanities Research Competition
- NAICS
- Scientific Research and Development Services (5417), Research and Development in the Physical, Engineering, and Life Sciences (except Nanotechnology and Biotechnology) (541715)
- SIC
- Services-Commercial Physical & Biological Research (8731), Services-Testing Laboratories (8734)
- akta.pro primary industry
- STEM Assessment, Badging & Competition Platforms (EDAFANAJ)
- akta.pro secondary industries
- Academic Olympiads & Subject Competitions (Science, Math, Informatics) (EDAMAGAE), Competitive Programming & Algorithms (DSA, Olympiad-Style) (EDAMACAI)
Keywords
Vesuvius Challenge business model
Business model- GTM type
- B2B
- Offering type
- Services
- Cost components
- Personnel, Technology or R&D, Operations
Revenue model
- Prize-based funding (donations allocated to prize awards): Vesuvius Challenge is a nonprofit funded by sponsors. All donated funds are distributed as prizes to participants who achieve breakthroughs in reading the scrolls. Over $1,800,500 has been awarded to date through competitions including Grand Prizes, monthly progress prizes, first-letters prizes, ink detection prizes, and open-source prizes.
Go-to-market motion2 records
Distribution channels2 records
Marketing channels6 records
Vesuvius Challenge product offering
Product offeringCore offering
Vesuvius Challenge is a machine learning, computer vision, and geometry research competition that reads carbonized Herculaneum papyrus scrolls without physical unrolling. It offers prize money to volunteer researchers who solve sub-problems (scanning, 3D segmentation, mesh flattening, ink detection) along a virtual unwrapping pipeline, and distributes open-source tools, ML models, and curated X-ray datasets to enable participation. All $1,800,500+ awarded to date has gone to teams producing breakthroughs such as the 2023 Grand Prize ($850,000) and the first full end-to-end scroll read (PHerc. 1667) in June 2026.
Product overview
Vesuvius Challenge is a machine learning and computer vision competition for reading carbonized Herculaneum scrolls. The portfolio consists of: (1) core segmentation tools including Volume Cartographer (manual/semi-automated virtual unwrapping), Khartes (interactive segmentation viewer), and Thaumato Anakalyptor (fully automated pipeline); (2) data access infrastructure via Python/C libraries and vesuvius-gui for browsing segments; (3) the Kaggle competition platform hosting the ink detection challenge; and (4) curated fragment datasets with ground truth ink labels for ML training. The pipeline combines 3D X-ray tomography, semantic segmentation, geometric reconstruction, and ink detection to virtually unwrap and read scrolls sealed since 79 AD.
Differentiator
Problem solved
Functional benefit
Products and services
- Volume Cartographer Open-source toolkit that lets researchers create meshes along manuscript surfaces, sample voxels around these meshes, and flatten them into 2D images. Includes a GUI for manual segmentation and command-line tools for automated rendering and processing, supporting the virtual unwrapping pipeline.
- Khartes Interactive manual segmentation tool for full-scroll movement through OME-Zarr data, providing live visualization of segmentation results. Pairs with Volume Cartographer for the flattening pipeline and runs on minimal hardware.
- ThaumatoAnakalyptor Fully automated 3D segmentation pipeline that combines classical Sobel kernel edge detection with Mask3D deep neural networks for point cloud instance segmentation, patch stitching via Monte Carlo methods, and SLIM-based UV flattening — handling surface detection and mesh reconstruction entirely automatically.
- Scroll Data Access Libraries (vesuvius) Python and C libraries providing programmatic access to Vesuvius Challenge scroll volume data without managing download scripts or storing terabytes of CT scans locally. Researchers can stream and process scroll data directly for ML training and analysis.
- Vesuvius Ink Detection Kaggle Competition Kaggle-hosted machine learning competition for detecting carbon-based ink in 3D X-ray volumes of Herculaneum papyrus fragments, providing dataset access, Jupyter notebook environments, and a competitive leaderboard.
- Curated Vesuvius Datasets Organized datasets for ML training, including manually annotated fiber skeletons, volumetric instance labels for semantic segmentation, and grand-prize banner region data with scan volumes and segmented meshes.
- Herculaneum Fragment Datasets Released datasets from synchrotron-scanned Herculaneum fragments — 6 fragments from Diamond Light Source (UK, 2023) and 3 fragments from ESRF BM18 (France, May 2025) — containing 3D X-ray volumes at resolutions from 2.2µm to 8.64µm, multispectral photographs in 16 bands (420-1050nm), and hand-labeled ink masks used for training supervised machine learning ink detection models.
Quantifiable outcome
- First entire Herculaneum scroll (PHerc. 1667) virtually unwrapped and read end-to-end, sealed since 79 AD (June 2026)
- +3 more outcomes
Companies that use Vesuvius Challenge
Customer profileSegments2 records
Ideal customer profiles2 records
Vesuvius Challenge technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- No
- API docs
- API detail
Core technology
AI maturity
App detail
Integration1 record
AI capability8 records
Feature6 records
Vesuvius Challenge partnerships and signals
Strategic signalPartnerships
Seven partnerships are on record, tiered core and major.
- European Synchrotron Radiation Facility (ESRF), GrenoblecoreEuropean synchrotron in Grenoble used BM18 beamline to scan three additional fragments in May 2025 at ultrafine 2.2µm voxel resolution using phase-contrast helical CT. Produced the highest-resolution scroll fragment scans to date.
- KagglemajorHosted the Vesuvius Challenge Ink Detection competition on Kaggle in 2023, which attracted thousands of ML practitioners and generated significant community engagement and model development for ink detection.
- Diamond Light Source (UK)coreUK national synchrotron science facility that scanned six Herculaneum fragments (Fragments 1-6) in 2023 using phase-contrast micro-CT at multiple resolutions and beam energies (3.24µm, 54keV and 88keV). Total scan data exceeding 600GB.
- EduceLab (University of Kentucky)coreEduceLab, led by Dr. Brent Seales, is the primary technical partner providing X-ray tomography scanning capabilities, Volume Cartographer software, and research infrastructure. The EduceLab team includes Dr. Brent Seales (Principal Investigator), Seth Parker (Research Manager), Christy Chapman (Research & Partnership Manager), and others who coordinate scanning operations and technical development.
- The Institut de France (Bibliothèque nationale de France)coreOwner/custodian of the Herculaneum papyri fragments. Provides physical access to the scrolls and fragments for scanning and maintains the collection. Françoise Bérard (Director of the Library at Institut de France) involved in fragment scanning operations.
- GettymajorGetty Museum supports the Vesuvius Challenge initiative as a partner organization, contributing to funding and institutional credibility.
- University of Naples Federico IIcoreItalian university providing the papyrology team including Federica Nicolardi (Lead Papyrologist, Assistant Professor), Marzia D'Angelo (Postdoctoral Fellow), Alessia Lavorante (Postdoctoral Fellow), Maria Chiara Robustelli (Postdoctoral Fellow), and Claudio Vergara (Postdoctoral Fellow) who evaluate prize submissions and transcribe recovered text.
Scale indicators5 records
Recent moves6 records
Expansion highlights6 records
Vesuvius Challenge competitors and assessment
Company assessmentDirect peers
- Topcoder: Open competitive programming and algorithm competition platform with cash prizes. Comparable to Vesuvius Challenge in that both host technical competitions with prize purses and rely on community-contributed solutions to hard problems.
- HeroX: Crowdsourced prize-based competition platform (founded by XPRIZE alumni). Comparable in that both run incentive-prize challenges to solve hard problems, distribute awards to individuals/teams, and use community participation as the primary delivery model.
- InnoCentive (now Wazoku): Open innovation and prize-based challenge platform that connects organizations to global solver networks. Directly comparable to Vesuvius Challenge in the prize-plus-crowdsourced-solutions business model, though applied to corporate R&D problems rather than research/heritage.
- Kaggle: Hosts the Vesuvius Challenge Ink Detection competition (2023). Kaggle is the most direct peer — both run ML/data science competitions with cash prizes, attract thousands of practitioners, and distribute open datasets and reproducible notebooks. The platforms operate at very different scales but share the competition-plus-community model.
- DrivenData: Hosts data science competitions for social impact, often in collaboration with nonprofits and research institutions. Directly comparable to Vesuvius Challenge in that both organize ML competitions with prize purses aimed at advancing research and humanitarian/heritage outcomes.
Broad incumbents
- DeepMind: World-leading AI research lab (Alphabet subsidiary) that applies ML to scientific problems (AlphaFold, etc.). Comparable as an institution applying cutting-edge ML to fundamental scientific problems, though at vastly greater scale, with commercial backing rather than nonprofit sponsorship.
- XPRIZE: Operates large-scale multi-million-dollar incentive prize competitions (Ansari XPRIZE, etc.). Comparable as the established incumbent of the prize-competition model Vesuvius Challenge operates in, though XPRIZE prizes are orders of magnitude larger and span many domains.
Others
- EduceLab (University of Kentucky): Brent Seales' academic research lab at University of Kentucky and primary technical partner of Vesuvius Challenge. Comparable as the originating research group that pioneered virtual unwrapping and provides the scan data, Volume Cartographer software, and pipeline R&D.
- Allen Institute for AI (AI2): Nonprofit AI research institute founded by Paul Allen that pursues ambitious research with philanthropic funding. Comparable to Vesuvius Challenge as a tech-billionaire-funded nonprofit advancing frontier AI/computer vision research in the public interest rather than for commercial return.
Emerging players
- CodaLab Worksheets: Open-source platform for running data science competitions and reproducible computational research. Comparable to Vesuvius Challenge's competition infrastructure as both enable reproducible ML experiment submissions, though CodaLab is academic-rooted and self-hosted.
Market position
Strengths4 records
Weaknesses4 records
Competitive moat5 records
Key risks6 records
Key highlights7 records
Customer concentration
Vesuvius Challenge social profiles
Digital presenceVesuvius Challenge financial estimates
Financial estimateRevenue estimate
Valuation estimate
Vesuvius Challenge leadership team
Management profileNumber of profiles
Profiles16 records
Vesuvius Challenge funding detail
Funding detailFunding overview
Funding rounds
Investors
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
Vesuvius Challenge M&A and investment
M&A and investmentM&A
Investments
M&A and investment is available on the Subscription and Enterprise plan.Contact sales →
Frequently asked questions about Vesuvius Challenge
What does Vesuvius Challenge do?
Vesuvius Challenge is a machine learning, computer vision, and geometry research competition that reads carbonized Herculaneum papyrus scrolls without physical unrolling. It offers prize money to volunteer researchers who solve sub-problems (scanning, 3D segmentation, mesh flattening, ink detection) along a virtual unwrapping pipeline, and distributes open-source tools, ML models, and curated X-ray datasets to enable participation. All $1,800,500+ awarded to date has gone to teams producing breakthroughs such as the 2023 Grand Prize ($850,000) and the first full end-to-end scroll read (PHerc. 1667) in June 2026.
Is Vesuvius Challenge a public or private company?
Vesuvius Challenge is a private company. It is classified as nonprofit foundation owned and is currently operating.
When was Vesuvius Challenge founded?
Vesuvius Challenge was founded in 2023. It employs 11 to 50 people.
How does Vesuvius Challenge make money?
One revenue line is on record: prize-based funding (donations allocated to prize awards).
Who are Vesuvius Challenge's main competitors?
Direct peers on record are Topcoder, HeroX, InnoCentive (now Wazoku), Kaggle and DrivenData. Broad incumbents are DeepMind and XPRIZE. Others are EduceLab (University of Kentucky) and Allen Institute for AI (AI2). CodaLab Worksheets is listed as an emerging player.
Does Vesuvius Challenge have an API?
No public API is recorded for Vesuvius Challenge.
What industry is Vesuvius Challenge in?
Vesuvius Challenge's product category is Digital Humanities Research Competition. Its primary akta.pro industry code is EDAFANAJ, STEM Assessment, Badging & Competition Platforms, with a secondary code of EDAMAGAE, Academic Olympiads & Subject Competitions (Science, Math, Informatics). Its NAICS code is 5417 and its SIC code is 8731.