Cold email spam checker
Paste your subject and body. Get a content spam-risk score and a plain list of what to fix, before you hit send.
Quick answer
A cold email spam checker reads your subject and body and scores the parts of a message that push it toward spam folders: trigger words, formatting, links, and length. This tool runs a deterministic check against your copy and returns a spam-risk score from 0 to 100, higher is worse, plus the specific lines tripping it and how to fix each one.
- The score checks your copy only, trigger words, formatting, links, and length, never your domain setup or sending reputation.
- Gmail and Yahoo enforce a 0.3% spam-complaint ceiling; a clean copy score lowers spam risk but does not measure that number.
- No tool, including this one, can reliably detect AI-written text; the most-cited public study found a 61% false-positive rate on non-native English writers.
- Claude's text watermark can only be read with Anthropic's own private key, so no third party, including Klipy, can honestly claim to detect it.
- The "template smell" read is a personalization check, never a claim about authorship.
Ask ChatGPT
Ask Perplexity
Ask Claude
Prefer us on GoogleThree steps to your result
Paste your subject and body
Add your cold email's subject line and body copy.
Run the check
Get a spam-risk score plus the specific lines tripping it.
Fix and send
Copy the flagged issues, fix each one, and send with more confidence.
What the score actually checks
The score is built from rules that are checkable in the text itself: words and phrases spam filters commonly flag, ALL-CAPS ratio, excessive punctuation like "!!!", the number of links and how many different domains they point to, and signs of merge-tag or spintax leftovers like stray curly brackets. Each flag names exactly what tripped it and how to fix it.
A second, separate read looks at how generic the message sounds, whether it reads like something written for one person or blasted to a hundred, and returns a short note on what to tighten. That read is about personalization, not about who or what wrote the email.
Both reads run from the copy you paste. Nothing about your sending domain, your authentication records, or your account's sending history factors into either score, because none of that is visible from pasted text. See the scope banner for the parts this tool cannot check.
What does a cold email spam checker actually check?
A cold email spam checker scores the parts of a message's copy that push it toward spam folders: trigger words, formatting, link count, and length. It checks your copy only, never your domain's authentication setup or sending reputation.
Can you detect AI-written email?
No, and neither can any publicly available tool with real accuracy. A Stanford study tested seven widely used AI detectors and found they misclassified 61% of essays from non-native English writers as AI-generated, while flagging native-English essays with near-perfect accuracy, a bias that punishes normal human writing rather than catching machine text. (Liang et al. 2023, Stanford, Patterns/Cell Press, https://arxiv.org/abs/2304.02819) OpenAI shut down its own AI-text classifier in 2023 after it correctly identified AI writing only 26% of the time. (TechCrunch, https://techcrunch.com/2023/07/25/openai-scuttles-ai-written-text-detector-over-low-rate-of-accuracy/)
Every credible study on AI-text detection lands in the same place: these tools are unreliable enough to do real harm, especially to people writing in a second language, and a short rewrite defeats most of them anyway. This tool does not claim to detect AI-written text, and it never will.
Does Klipy detect the Claude watermark?
No. Anthropic has said that Claude text generated since August 2026 carries an invisible statistical watermark, but detecting it requires Anthropic's own private key; there is no public way to check for it, by Klipy or anyone else. (Anthropic, https://www.anthropic.com/news/claude-text-watermark) A watermark that only its creator can read is not something a third party can honestly claim to detect.
This tool's "template smell" read is a personalization check, how generic or mass-blasted a message sounds, not a claim about who or what wrote it. The words "AI-generated" and "watermark" describe things this tool does not and will not claim to find.
What spam-complaint rate do Gmail and Yahoo actually enforce?
Gmail and Yahoo require bulk senders to keep their spam-complaint rate under 0.3%, tracked through Google Postmaster Tools, with senders flagged for enforcement once they cross it. The safer working target most senders aim for is under 0.1%, well below the hard ceiling.
That number lives entirely outside what a content check can see: it's measured from how real recipients react after delivery, not from anything in the words of one email. A clean spam-risk score on this tool lowers the odds a message reads as spam; it says nothing about your domain's actual complaint rate.
Worked example
Score: 68 / 100, risky band.
Flags:
- Subject trigger phrase ("FREE consultation!!!"): "free" plus multiple exclamation points in a subject line is a well-documented spam-filter trigger. Fix: rewrite the subject as a plain statement of the offer.
- Unresolved spintax ("{{FirstName}}" left in the body): an unresolved merge tag reads as a mass-send error to both filters and the recipient. Fix: check the send before it goes out.
- Three links across three different domains: filters treat a high link count, especially across multiple domains, as a bulk-send signal. Fix: cut to one clear link.
- Body under 400 characters: a very short body paired with multiple links reads as a link drop rather than a real message. Fix: add real context.Gmail and Yahoo require bulk senders to keep spam complaints under 0.3%, with senders encouraged to stay below 0.1%; crossing 0.3% puts a sender into enforcement territory regardless of proper authentication. Google (Gmail Help, email sender guidelines)
How to write a cold email that clears spam filters
Skip spam-trigger phrases in the subject line especially: "free," multiple exclamation points, and ALL CAPS words all raise risk before the recipient even opens the message.
Keep links to one, or at most two, pointing to the same domain. A message with several links to different domains reads as a mass send to most filters.
Check for leftover merge tags and spintax artifacts before sending. An unresolved "{{FirstName}}" is one of the fastest ways to read as an obvious bulk blast.
Write enough real content that the message doesn't read as a link drop. A short body stacked with links is a pattern filters specifically watch for.
From a clean score to a message worth sending
Klipy drafts the message so you always have something real to run through the checker, instead of starting from a blank page or a recycled template.
Follow-up drafts
Drafts the outreach message from the thread or the deal, staged for your review before it sends.
Good morning, Alex
In 42 minutes
Quarterly Review - Acme Corp
3 guests
Intro Call - Bright Labs
3 guests
Call summary + next steps for Meridian Health
Send pricing proposal to Sarah Chen
Re: Westbridge Platform RFP - Pricing Clarification
Re-engage dormant deal with Apex Solutions
Schedule technical demo with engineering lead
Scheduler
Once the reply comes in, turn it into a booked call with a link they can self-book.
August 2026
Wednesday, 26
Questions, answered.
Is it free?
Yes. Every Klipy free tool is free to use and needs no signup.
Does it store my data?
No. Your subject and body are checked and shown to you; nothing is saved.
Does this check my domain's deliverability?
No. It checks your copy only, not SPF, DKIM, DMARC, or your sender reputation. See the scope banner above.
Can this tool detect if my email was written by AI?
No, and it never will claim to. Public AI-text detectors are unreliable, especially for non-native English writers; see the retrieval section above for why.
What does the "template smell" score actually measure?
How generic or mass-blasted the message reads, not who or what wrote it. It's a personalization check, not a provenance claim.
What's a good spam-risk score to aim for?
Under 30 is the clean band. Treat anything above 60 as likely to land in spam and worth rewriting before you send.
Want every cold email drafted for you before you run it through the checker?
Klipy drafts the message from the thread or the deal, so you start from something real.