Does This Read Like ChatGPT? Preview Before Submission
Worried the draft that you worry may resemble ChatGPT output may be flagged? Preview how the analyzed text reads to OmniDetect, inspect the passages that reacted, revise for meaning, evidence, and language you can explain, then recheck the new version.
This preview cannot determine authorship, reproduce the teacher, employer, client, or publisher's checker, or replace the recipient’s policy and review process.
Preview my documentWhat the checker-view preview shows
A bounded view of the analyzed excerpt—not a decision by the recipient.
Observed result
The result band and passages that reacted within the text OmniDetect actually analyzed.
Revision target
Places to inspect for meaning, evidence, and language you can explain; never an instruction to disguise the writing process.
Decision boundary
The teacher, employer, client, or publisher can use a different method and applies the recipient’s policy and review process.
Free preview: we analyze up to the first 2,000 characters
The free preview analyzes up to the first 2,000 characters. Longer text remains available if you continue after signing in.
We keep your document so you can find it again, and you can delete it whenever you want. We do not sell it. Manage or delete
Preview → revise for meaning → recheck
Preview the submitted version
Paste the same draft you plan to send and confirm the analyzed scope before interpreting the result.
Revise what needs real improvement
Review the reacted passages for meaning, evidence, and language you can explain. Preserve correct material instead of chasing a score.
Recheck the revision
Compare the new OmniDetect result. A changed signal is useful feedback from this method, not a promise about another checker.
Pre-submission review checklist
- Read the recipient’s policy and review process before submitting.
- Confirm the analyzed range and method boundary shown with the result.
- Review passages for meaning, evidence, and language you can explain.
- Keep factual claims, quotations, citations, names, and dates correct.
- Preserve normal drafts and version history as part of the writing record.
- Do not change correct writing solely to imitate a supposedly “human” style.
Evidence and limits behind this workflow
Reviewed 2026-08-27Turnitin Guides
Its current guide says the model can misidentify human and AI-written text and should not be the sole basis for adverse action.
Scope: Turnitin product guidance for qualifying submission prose; it does not validate OmniDetect or describe every recipient’s process.
Liang et al., Patterns (2023)
The study found substantial false-positive risk for 91 human-written TOEFL essays across seven detectors.
Scope: A small, specific non-native-English essay dataset; evidence of risk, not a universal rate for this page or OmniDetect.
Continue with reviewed context
How to read detector claims
Check dataset scope, false positives, analyzed range, and validation status before trusting a number.
Current OmniDetect benchmark status
See what has been measured, what is historical, and which current accuracy numbers remain withheld.
Already received a flag?
Preserve version history, drafts, and process evidence; another detector score does not decide the case.
ChatGPT concern revision-kit pilot
Planned pilot: selected users whose eligible free scan returns confident-AI Red may see a $9.99 one-time option for passage-specific guidance and rechecks for up to 7 days.
- One document, up to 10,000 characters
- Guidance tied to the passages observed by OmniDetect
- Meaning-focused revision and repeat rechecks for up to 7 days
- No authorship ruling and no promise about the teacher, employer, client, or publisher's checker
Pilot checkout is not live. The slot is shown here for validation only.