Use caseFor editors and content teams
Fact-check the draft before it goes out.
For articles, newsletters, briefs and posts, whoever wrote them, Lenz identifies the factual statements it can check against independent public sources and returns a result for each. It also checks the draft's citations: fake DOIs, dead links, and sources that don't say what the sentence says.
Example: a newsletter paragraph“Spotify reported 696 million monthly active users in the second quarter of 2025, up 11% from a year earlier. The company was founded in Stockholm in 2006 and launched in the United States in July 2009.”
FalseSpotify launched in the United States on July 14, 2011.
- A result per statementnot one verdict for the document
- Citations checked toofake DOIs, dead links, misquoted sources
- Warranted verdictson qualifying checks, on Pro and Scale
01Example
Three statements. One date is wrong.
| Claim | Verdict | Lenz Score | Sources |
|---|---|---|---|
| Spotify reported 696 million monthly active users in the second quarter of 2025, an 11% increase from the second quarter of 2024. | Truehigh confidence | 10/10 | sec.govnewsroom.spotify.coms29.q4cdn.com+14 more |
| Spotify was founded in Stockholm in 2006. | Truehigh confidence | 10/10 | allabolag.sesec.govlei.report+15 more |
| Spotify launched in the United States in July 2009.Suggested rewriteSpotify launched in the United States in July 2011. | Falsehigh confidence | 1/10 | techcrunch.comtheverge.comengadget.com+15 more |
Deep check · Checked 7 September 2026.
Open a row for the sources it used and the reasoning that weighed them.
02How it fits your desk
Lenz checks the facts.
The editorial call stays yours.
Automated fact-checking for articles, newsletters, briefs, posts and reports before publishing, whoever wrote them.
-
The draft is ready
An issue, a brief, a post, a report: whether a model wrote the first pass or you did.
-
Lenz finds the checkable statements
You hand over the whole piece rather than a list. Lenz picks out the factual statements it can check against independent public sources and leaves the opinions and the recommendations alone. It also checks the draft's citations: whether each source says what the draft says it does, with the passage it rests on.
-
Each one comes back with a result
A quick check returns a verdict and a confidence level per statement in about fifteen seconds; that is how you sort a whole issue. Send the ones that matter to a deep check, where the verdict comes from sources Lenz went and found.
-
You decide what to publish
Read the results before the issue goes out: cut a line, fix a figure, or run a deep check on the one you are least sure of. The editorial call is yours.
-
Publish, with the checks behind it
Each deep check keeps a record you can reopen: the statement as framed, the verdict and score, every source with its date, the reasoning. A figure questioned later has a check behind it.
03Citations
Catch the source that doesn't say what the sentence says.
Verify a draft's citations for fake, made-up and hallucinated references.
Lenz reads the links and DOIs in the draft and flags a DOI that is not registered or does not match the reference, a page that is gone, and a quote the page does not contain.
It also flags a source that doesn't say what the sentence says, with the passage beside it. When a source supports only part of the sentence, the citation is marked Needs a closer look, with the passage the source does give.
A reference with no link and no DOI is not checked.
04What editors check
Anything with a figure, a date or a name in it.
-
Market briefs, news digests and articles
Paste the whole piece. Lenz returns a result per statement, not one verdict for the document, so a wrong figure is named with its own sources.
-
Blog posts, press releases and social media posts: the statistics in them
A statistic is checked like any other statement: Lenz looks for where the number was published and reports what it found, including when that is nothing.
-
A writer's draft, before you sign off
A writer's draft comes in. Paste it, and Lenz lists its checkable statements with a first read on each, so the ones worth a second look stand out before you approve it.
05In your document
Get the draft back in Word, with comments and suggested edits.
The company was founded in Stockholm in 2006 and launched in the United States in July 2009July 2011.
Word and Google Docs
In the Workbench on a desktop, upload the draft as a Word document (.docx), or paste a link to a Google Doc shared with anyone who has the link, there or on lenz.io/verify. You download the same file with a comment on each claim and citation that needs attention.
Suggested edits as tracked changes
Accept or reject each one in Word. In the Workbench, “Your draft” shows the wrong part struck through with the replacement beside it. Not every claim gets one, and a suggested edit is not itself verified. Lenz does not write back to the Google Doc.
06Where you work
In Claude or ChatGPT.
Check LLM output, AI documents, ChatGPT-generated content and X or LinkedIn posts in Claude or ChatGPT.
- 01Connect LenzOne sign-in, no key.
- 02Paste the post or the draft, and ask“Check the claims in this draft with Lenz.”
- 03Read the results in the same conversationA result for each statement; ask for a deep check on the one you are least sure of.
- Spotify reported 696 million monthly active users in the second quarter of 2025, up 11% from a year earlier.Truehigh confidence
- Spotify was founded in Stockholm in 2006.Truehigh confidence
- Spotify launched in the United States in July 2009.Falsehigh confidence
Drawn from the example above. Claude and ChatGPT each lay out their own card.
07Automate it
Run the check as part of how an issue goes out.
If a workflow already prepares your newsletter or your posts, add the check as a step in it. Each statement comes back with its own result, the same as in the Workbench.
- n8nAdd a Lenz step to a workflow that already prepares your issue or your posts. The node is on n8n.io.
- ZapierUse Lenz as a step in a Zap, with no code: send the draft in and pass the results on to the next step.
- A scriptSend each draft from your own script, in Python or TypeScript, and read the results before it goes out.
- A coding agentConnect Lenz to Claude Code, Cursor, VS Code or Codex, sign in, and ask it to check a draft in your project.
08Deep check
The investigation, not a longer quick check.
- 01Framingprecise enough to test
- 02Researchindependent public sources
- 03Debateboth sides argued
- 04Panel Reviewthree reviewers
- 05Conclusionverdict, sources and reasoning
Lenz searches independent public sources for the statement, has models argue both sides on what it found and a panel review the arguments, then returns the verdict with those sources and the reasoning.
About a minute or two per statement.
Warranty
Vendors disclaim accuracy. Lenz warrants the verdict.
A verdict can qualify when a full verification returns True or False with high panel confidence and enough independent sources behind it.
On a paid Pro or Scale plan, every full verification that qualifies carries a contractual warranty, at no extra charge.
09Pricing
Try free, no card.
$7.99/ month
500 quick checks or 50 deep checks a month, in any combination.
Or $6.59 a month, billed $79 a year
Choose Plus$99/ month
5,000 quick checks or 500 deep checks a month, in any combination. The warranty on qualifying verdicts.
Or $83 a month, billed $990 a year
Choose Pro$399/ month
20,000 quick checks or 2,000 deep checks a month, for production volumes. The warranty on qualifying verdicts.
Or $333 a month, billed $3,990 a year
Choose Scale10Questions
No. Connecting Lenz to Claude is one sign-in and needs no key. Cursor and Claude Code take one command or one click, then the same sign-in. You can also use a script to automate checks.
Two different checks. The quick check is a first read: a panel of models reads the statement and returns a verdict and a confidence level in about fifteen seconds. The deep check is an investigation: Lenz searches independent public sources and returns the verdict with those sources and the reasoning, in about a minute or two per statement. The deep check can change the quick verdict. When it does, rely on the deep check: it is the one that shows its sources and its reasoning.
Nothing is invented to fill the gap. When no public record settles a statement, Lenz answers “No Public Record” instead of a score. That doesn’t make the statement false.
Drafts you check are private to your account, whether you paste them on the site, check them in Claude or ChatGPT, or run the command line. Anthropic, Google, OpenAI, Perplexity and xAI, the AI model providers Lenz uses, do not train their models on what Lenz sends them. By default Lenz keeps your drafts and their results while your account exists; on Pro and Scale you can set a shorter period, including zero, where what you submit is never written to Lenz’s database. The model providers keep what their own terms say, listed on the subprocessors page.
Yes, in the Workbench on a desktop: upload a .docx, or paste a link to a Google Doc shared with anyone who has the link. A Google Doc link also works on lenz.io/verify. You download the checked file as a Word document with Lenz’s comments and suggested edits. Lenz does not write back to the Google Doc.
Not answered here? Talk to us