Licensing

Wikexa is a processed, indexed and ranked knowledge service. Using it is free at ordinary volumes. Commercial volume, model training, bulk access and partnerships are licensed — talk to us.

What you are licensing

The service, the index, and everything behind it.

Wikexa is not a file server. What makes it worth calling is work we built and maintain:

  • The ingest and extraction pipeline — how raw source material becomes structured records, which fields survive, how templates and references are resolved, how entries are shaped per language and part of speech.
  • The search index and ranking model, including popularity weighting, exact-match bonuses and disambiguation penalties.
  • The alias and entity resolution indexes that let any spelling, redirect or identifier land on the right record.
  • The compilation — the selection, structure and arrangement of the whole corpus — and the database rights in it.
  • The derived metadata: relevance scores, popularity signals, snippets, cross-references and structured summaries.
  • The API, schemas and software, and the Wikexa name and marks.

All of it is ours, protected by copyright, database, trade secret and other laws, and reserved. Access is governed by our Terms of Service.

Access

Free to use. Licensed to exploit.

Free
$0

Individuals, research, education, open-source projects, and ordinary production use by applications and agents.

  • Every tool, whole corpus
  • No key, no account, no signup
  • Fair-use rate limits
  • Best effort, no SLA
Commercial & partnership
Talk to us

Sustained volume, model training rights, bulk or snapshot delivery, private deployment, or a commercial partnership.

  • Negotiated throughput and SLA
  • Bulk export / private instance
  • Training and redistribution rights
  • sales@wikexa.com

A licence is required if you are training models on our output, extracting the corpus in bulk, redistributing or reselling access, or running at a volume that looks like a data acquisition programme rather than an application. We would much rather agree terms up front than discover it in our logs.

For AI labs and platforms

Partnership beats scraping, for both of us.

If you are building models or agent platforms at scale, crawling us is the worst version of this deal: you get rate-limited output, no freshness guarantee, no bulk delivery, and no recourse when a schema changes.

The alternative is a straightforward one — negotiated bulk or snapshot access, guaranteed refresh cadence, custom record shapes, coverage extended to whichever corpora you actually need, and a contract behind it. We are a small team, we move quickly, and we are open to anything from a data licence to a deeper partnership.

sales@wikexa.com

Source material

Third-party content and attribution.

Wikexa indexes material that originates with third parties, each source under its own licence. We make no proprietary claim to that source material: the rights described above subsist in our own work, not in it.

Every record we return carries the source_url and license for the material behind it. If you republish that material, meeting the terms of its licence — attribution, a link back, share-alike, or whatever else it requires — is your responsibility:

"source_url": "https://example.org/the-original-page",
"license":    "CC BY-SA 4.0"

Licences vary by source. Some are public domain and impose no conditions at all; others require credit and a link. Read the license field rather than assuming. Facts alone — dates, quantities, relationships — are not copyrightable and carry no such obligation whatever the source.