api.originality.ai/api/v1/scan/ai: POST + X-OAI-API-KEY Guide
If you searched for api.originality.ai/api/v1/scan/ai or X-OAI-API-KEY, use this page as a practical integration checklist before wiring AI detection into an app, CMS, LMS, or editorial workflow.
Quick answer
What developers need to know about api.originality.ai/api/v1/scan/ai
Originality.ai publishes API documentation for automated AI and plagiarism scans, and its help center points developers to the current API docs for endpoint and authentication details. Treat api.originality.ai/api/v1/scan/ai as a commonly searched v1-style endpoint reference: verify the active docs, keep keys server-side, test response formats, and never let an AI score become the only evidence in a high-stakes decision.
Search-result answer: use a backend POST request to https://api.originality.ai/api/v1/scan/ai, send X-OAI-API-KEY as a server secret, and treat lowercase x-oai-api-key in logs as normal header normalization unless authentication fails.
If your search includes scan/ai, start with this endpoint checklist
Searches for api.originality.ai/api/v1/scan/ai usually come from developers debugging a v1-style integration, a copied snippet, or a connector reference. Before changing code, confirm these four details against the current vendor docs:
- 1. Base URL and route: check whether your account should use the current docs route, the legacy v1 route, or a newer scan path.
- 2. Method and payload: AI content scans are submitted as API requests with text content; URL scans use a separate webpage-scan flow.
- 3. Header: keep
X-OAI-API-KEYin server-side code only. - 4. Confidence policy: use word count and review thresholds before acting on
score.aior similar response fields.
Reviewed June 2, 2026
Current docs snapshot: v1 scan/ai, newer scan docs, and X-OAI-API-KEY
The safest interpretation is practical: api.originality.ai/api/v1/scan/ai is a recognizable v1-style endpoint search, but production integrations should verify the current docs path, account permissions, and response shape before deployment. Source links were rechecked June 2, 2026 against the provider help center, API docs, short-text note, and Microsoft connector reference. If your issue is a failing cURL request, lowercase x-oai-api-key log, or invalid API call error, use the focused scan/ai troubleshooting guide.
Citation routing: cite this endpoint checklist for the exact api.originality.ai/api/v1/scan/ai question; cite the blog troubleshooting guide when the user needs cURL examples, invalid API call debugging, or provider comparison.
Official API entry point
Originality.ai still describes a REST API for automated AI and plagiarism scans and links users to the official API documentation.
Integration move: Start from the vendor docs before copying old endpoint snippets from forums or connectors.
Version drift
The public docs navigation now exposes newer scan flows such as POST Scan, POST Batch Scan, POST Scan Url, GET Credit Balance, and GET Scan Results.
Integration move: Treat v1 scan/ai examples as legacy-compatible references until your account docs confirm the exact production route.
Short text risk
Originality.ai says API AI scans have no word-count minimum, but accuracy is reduced below 100 words.
Integration move: Label chat messages, titles, captions, and short emails as low-confidence even when a numeric score is returned.
Authentication and fields
Microsoft connector docs expose X-OAI-API-KEY as a secure string and show response families such as success, score.ai, score.original, credits, and word count.
Integration move: Keep keys server-side, handle failures separately from scores, and log billing metadata without storing private text unnecessarily.
| Integration area | What to verify | Why it matters |
|---|---|---|
| Endpoint version | Confirm whether your integration uses the current docs, legacy v1 routes, or a newer scan flow. | Copying an old snippet can fail silently or return a response shape your app does not handle. |
| Authentication | Keep API keys server-side and pass them through a backend proxy, not from client JavaScript. | A leaked X-OAI-API-KEY can burn credits, expose scan history, or create account risk. |
| Text length | Treat short text as low confidence; Originality.ai notes API scans have no minimum, but accuracy drops below 100 words. | Short emails, chats, captions, and titles are where false positives and false negatives are most likely. |
| Decision threshold | Set review bands such as low, manual review, and high risk instead of one hard fail score. | A single probability score can become an unfair disciplinary or editorial decision. |
| Audit trail | Store input length, model/version, timestamp, score, reviewer, and final action without over-retaining private text. | Without an audit trail, teams cannot explain a decision or handle appeals. |
Response fields worth handling before launch
| Field family | How to use it | Launch caution |
|---|---|---|
| success / status code | Confirm the request completed before trusting any score. | Retry, timeout, and billing failures should not be treated as human or AI results. |
| score.ai / score.original | Convert vendor probabilities into internal review bands such as low, manual review, and high risk. | Do not turn a probability into an authorship verdict without context and review. |
| credits_used / credits | Track cost per scan, user abuse, and monthly budget burn. | Log billing metadata separately from private text where possible. |
| word_count / wordCount | Flag short samples automatically because vendor docs warn accuracy drops below 100 words. | A short text score should be labeled low confidence even when the API returns a number. |
Microsoft's connector reference exposes AI detection operations and response field families such as success, score, credits, word count, and block-level results. Treat that as implementation context, not a substitute for the active vendor API docs.
Originality.ai API
Publisher workflows, team scanning, plagiarism plus AI checks, and high-volume text review.
Check: Verify the current docs, endpoint version, key header, retention settings, and billing before launch.
GPTZero / Copyleaks APIs
Education, LMS, and enterprise workflows where vendor support and reporting matter.
Check: Check policy fit, appeal handling, non-native English false-positive risk, and regional compliance.
EyeSift browser checks
Manual triage, privacy-first draft review, short-sample warnings, and false-positive interpretation.
Check: EyeSift is not a bulk API endpoint today; use it as a review tool, not a server integration.
Originality.ai API integration notes
Originality.ai says its API is designed to bring AI detection and plagiarism checks into publisher, education, agency, and platform workflows. Its help center points developers to official API docs for endpoints, authentication, and implementation details, including the current v1 documentation path.
Microsoft's connector documentation references X-OAI-API-KEY as a secure API key field for Originality.ai integrations. That is a strong hint that keys should stay in backend infrastructure, secret managers, or server-side environment variables.
Originality.ai also notes that API scans have no word-count minimum, but accuracy is reduced below 100 words. That matters for product design: a 40-word chat message, a title, or a short email should be labeled low confidence even when an API returns a numeric score.
Do not hard-fail users from one AI score
AI detection APIs output probabilities. For academic integrity, hiring, publishing, marketplace moderation, or legal workflows, route medium/high scores to human review, preserve process evidence, and provide an appeal path. This is especially important for short samples, non-native English writing, translated text, technical documentation, and heavily edited drafts.
When EyeSift is the better first step
If you are not building a bulk scan pipeline, a browser-based review can be safer and faster. EyeSift's text detector runs in the browser, shows confidence and false-positive warnings, and is useful for checking a draft before deciding whether a paid API integration is worth the complexity.
Need the full developer troubleshooting article? Use the Originality.ai scan/ai API guide for cURL examples, X-OAI-API-KEY handling, lowercase header logs, short-text confidence rules, and production review safeguards.
AI Text Detector
Private browser check with confidence, short-sample warnings, and human-writing signals.
AI Detector Appeal Letter
Build an evidence packet when a detector score might be a false positive.
Plagiarism Checker
Use exact-match and similarity checks alongside AI detection signals.
False-Positive Rate Guide
Understand why detector accuracy changes by content type, length, language, and context.
Sources and docs to verify before deploying
- Originality.ai API Help Center — overview of API use cases and link to official API documentation.
- Originality.ai API Documentation — current docs for scan endpoints and implementation details.
- Originality.ai Minimum Word Counts — API scans can run below 100 words, but accuracy is lower on short text.
- Microsoft Originality.ai Connector — connector reference that documents the secure API key field.
- FTC Guidance on AI Claims — reminder to avoid unsupported accuracy or automation claims.
Last reviewed June 2, 2026. API documentation can change; verify vendor docs before deploying production code.
Frequently Asked Questions
What is the Originality.ai AI detection API endpoint?
Originality.ai documentation and integrations reference AI scan endpoints under api.originality.ai, including v1 scan endpoints and newer documentation versions. Use the official Originality.ai API docs as the source of truth before deploying because API versions can change.
Is api.originality.ai/api/v1/scan/ai still the right endpoint in 2026?
Treat api.originality.ai/api/v1/scan/ai as a v1-style endpoint reference. Originality.ai still points API users to official documentation, while the public docs also show newer scan flows. Check the active docs for your account before changing production code.
What is X-OAI-API-KEY?
X-OAI-API-KEY is an API key header used by Originality.ai integrations to authenticate API requests. Keep it server-side and never expose it in browser JavaScript, public repositories, or static client bundles.
What fields should developers handle from an AI detection API response?
Handle request success or status fields, score fields such as AI and original probabilities, credit or billing fields, and word-count fields. Word count matters because short samples should be treated as lower confidence even when the API returns a numeric score.
Does EyeSift offer a public AI detection API?
EyeSift is currently a browser-based AI detection and writing-tool suite. It is useful for manual review, false-positive triage, and browser-only checks, but it should not be presented as a drop-in API replacement for paid bulk scanning APIs.
Can an AI detection API prove that text was AI generated?
No. AI detection APIs return probability signals, not proof of authorship. Use thresholds, manual review, audit trails, and an appeal path for academic, hiring, legal, or editorial decisions.