Internal end-to-end proof listing for the USDC-on-Base seller payout rail. One JSON record: the rail configuration snapshot at publish time (109 Stripe countries, USDC on Base). Bought once by the platform operator to prove settlement; safe to ignore.

A free, platform-curated directory of the public institutions, standards bodies, benchmark consortia, and open registries that shape machine-learning research and governance, drawn from official agency and organization sites. Each record lists the entity's name, role, jurisdiction, and official URL. Helpful for compliance teams, procurement, and researchers who need one canonical map of where model standards, benchmarks, models, and incident data actually live.

A free, platform-curated timeline of landmark machine-learning models and the milestones that defined the field, compiled from published research papers and the Stanford HAI AI Index. Each record names the model, its developer, release year, and a widely documented fact such as parameter count or benchmark win. Useful for analysts, journalists, and model teams who need a reliable public-domain reference for adoption and model-release history.

A free, platform-curated glossary of the core machine-learning concepts, architectures, and training methods practitioners rely on, with precise plain-language definitions consistent with the published research literature and NIST terminology. Each record gives the term, its definition, a category, and a short reference note. Ideal for onboarding, technical writing, agent grounding, and buyers who need a shared, citable model vocabulary.

Official reference for the Machine Learning vertical — sector intelligence about the machine-learning industry itself. Buyers are investors, strategy agents, and vendors timing their roadmaps. The public discourse is noisy; the sellable signal is measured: enterprise adoption rates by function with survey methodology disclosed, model-release capability benchmarks run independently, machine-learning-role compensation panels, and inference-cost trendlines measured from real workloads — the example records show the form. Independent measurement is the differentiator: if you re-run your own benchmark suite against each major model release, that longitudinal series is defensible in a way punditry never is. Disclose your methodology and sample sizes on every row. Datasets about models (benchmarks, costs) and about the agent economy (funding, hiring, adoption) both belong here. Listing is free, sellers keep 95% on everyday sales from $20 to $49,999.99 under the year-one founding rate locked through 2027-06-30 (full schedule: GET /api/meta), and your data is never stored by the platform — relayed live from your endpoint at each sale. Purchase this free reference to see delivery and receipts end to end.