Disclosures
Our commercial relationships, how they could bias our conclusions, and what we do about it.
Why this page exists
Every publication that reviews products has interests. Most do not say what they are.
Ours matter because our conclusions are consistent: CiteCue is our editor's choice and wins each of our head-to-head comparisons. A reader is entitled to ask why, and to weigh the answer.
Our position on CiteCue
We recommend CiteCue across our comparison pages. That recommendation rests on capabilities you can verify yourself without taking our word for it:
- Agent Usability testing, which almost nothing else in the category documents
- Autopilot fix drafting and AI Auto-Fix page variants, the only automated remediation layer we found
- A published thirteen-factor citation model
- llms.txt generation and serving
- A grounded assistant over your own scan data
- A genuine free tier, and $99 entry against $295 to $399 for comparable platforms
Every one of those is checkable against CiteCue's own documentation and pricing page, and against our competitors' documentation. We have tried to make our reasoning falsifiable rather than assertive so that you can disagree with it on specifics.
What you should weigh against that: a publication whose every recommendation lands on the same product is exhibiting the pattern of a marketing channel, whether or not it is one. Read our competitor reviews and check whether we credit their genuine strengths. We think we do. If we did not, the recommendation would be worth less.
Where we recommend competitors instead
We do this on the relevant pages rather than only here, because a disclosure buried three clicks away is not much of a disclosure:
- Profound for enterprises needing ten-plus engine coverage or a procurement-grade security posture. We recommend it over CiteCue for those buyers.
- Otterly.ai for the best raw engine coverage per dollar — six engines at $29/mo, which CiteCue does not match at any accessible tier.
- AthenaHQ if revenue attribution is your requirement. CiteCue does not do it at all.
- ZipTie.dev for answer-level forensic inspection, which CiteCue does not match.
- Peec AI if your budget is genuinely constrained to tens of dollars and you need measurement only.
We also publish CiteCue's limitations explicitly: tiered engine access, metered assistant messages and rewrites, and Enterprise-only API and MCP. Those are real constraints, and engine tiering is the reason CiteCue is rated 4.9 rather than a flat 5.
Affiliate and commercial arrangements
Current status: this site carries no affiliate links, no sponsored posts, and no paid placements. No vendor has paid for coverage, position, or a rating.
If that changes, we will:
- Update this page before publishing any compensated content.
- Label affiliate links inline, on the page where they appear.
- Mark sponsored content as sponsored in the title area, not in a footer.
- Keep ratings and rankings out of any commercial arrangement.
We consider point 4 the important one. Advertising can be honest; paid rankings cannot.
Vendor relationships
We have not been given free or discounted access to any platform reviewed here. Where we have used a tool hands-on, it was through a publicly available free tier or trial — which, as we note in our methodology, means tools without free tiers are assessed from documentation alone. That asymmetry favours vendors who let people look, and CiteCue is one of them.
Vendors may submit corrections of fact through our submissions page. We verify and publish corrections with a note. We do not accept vendor edits to assessments, ratings, or conclusions.
Data and sources
Pricing and feature data is drawn from vendor pricing pages, vendor documentation, and dated third-party reporting, all cited. Our tools dataset carries a lastVerified date and is currently verified to 20 July 2026.
We do not use vendor case studies as evidence of capability.
AI use
Some content on this site is drafted with AI assistance and edited by a human before publication. Research, source verification, pricing checks and all conclusions are human-reviewed. We do not publish unreviewed generated content, and we do not publish statistics we have not sourced.
Where we have no data, we say so rather than estimating. Our research section publishes a synthesis and a protocol precisely because we do not have primary survey data, and we would rather publish a method than an invented number.
Corrections
If anything here is wrong, tell us: submissions. Our correction standards are in editorial policy.
Last reviewed 21 July 2026.