Silvertext's checks are available to developers through a REST API, with an MCP server for agents. Both use the same key and draw from the same credit balance as the app.
Getting a key
Create a key under Account → API in the app. Keys are shown once, carry the scopes you pick (create checks and manage source rules, read reports and source rules, read usage), and can be revoked at any time. The first key on an account comes with starter credits.
What the API does
The same pipeline the app runs: submit text or a file, get a report with matched passages and their sources, language findings, citation findings and AI likelihood. The options field selects which checks run. Source rules (which web sources a plagiarism check may cite) can be managed for the account or sent with a single check, the same rules Account → Settings edits.
The MCP server
Agents that speak the Model Context Protocol can run checks, read reports, manage source rules and read the credit balance through the Silvertext MCP server, so a writing agent can check its own draft before handing it over. The MCP page on silvertext.com has the per-client setup.
Documentation
The developer docs at silvertext.com/developers carry the getting-started guide, guides and how-tos per language, and an endpoint reference generated from the live specification, so it always matches the deployed version.
Pricing
API usage is billed in the same credits as the app, at the same rates per page: 40 credits for every check plus each check the request turns on, so a plagiarism-only request (110 credits per 250 words) costs less than one with everything on (150). The options field on a request selects the checks; the response quotes the estimate, and the finished check reports what it cost. Department and volume arrangements are made directly; write to us from the contact page.