quietmoney

Audn

Runs black-box penetration tests and red-team exercises on voice AI, AI agents, and websites using autonomous attack agents, billed per scan or by monthly subscription. It also offers its own uncensored attack/defense model Pingu Unchained, a desktop command-line tool AudnCode, and an open-source adversarial learning framework.

audn.ai ↗Developer toolsJapan 90.05%, Brazil 9.95%#91 of 563 by revenue
The founder kept launching "offensive AI security" projects on Hacker News and Product Hunt. In 2026-08, new tiers such as the $99 one-off scan and the $999 subscription took monthly revenue from about $1,400 to $14,789 in one jump; the specific trigger event was not found.One-line growth story, compiled by us from public sources; speculation is marked separately below.
Open Audn ↗

Traffic model, last monthFrom first hearing about it to paying. Circles show how many passed through last month

Grey balls: target audience, not users yetGreen balls: Users (subscribed customers coming back to run scans, or people who kept browsing after arriving)One ball ≈ 336 people (100 balls every 5 seconds stand for one month)Click a circle for details and source links
Where the numbers in this chart come from

The main engine is viral spread (the founder's project launches and open source reputation): the top search keywords are all branded ones such as audn.ai download and audn.ai platform, which shows people hear of it elsewhere first and then search.

Hacker News and Product Hunt launches were only small bursts, and no turbo boost was found; last month's visits were up 275.3% month over month, but 90% of visitors are in Japan, and the reason was not found.

By our estimate, visit counts are estimated (from the global rank) and are probably too low; about 6.4 payments (including renewals) per 100 visits, far higher than a typical site.

Model: Lenny's racing growth framework. Engine = viral spread (the founder keeps launching projects on Hacker News, Product Hunt and X, and the publicly testable models and open source frameworks bring branded keyword searches); lubricant = monthly subscriptions and add-on packs; turbo boost = not found.

Real data: Direct being the largest channel at 30.91%, the channel ranking (direct, organic search, organic social), 17 search keywords all brand-related, countries (Japan 90.05%, Brazil 9.95%), bounce rate 71.61%, 1.15 pages per visit and month-over-month +275.3% come from the Similarweb public page, last month; the 4,000 total visits are not a real figure but are estimated from the Similarweb global rank #4529488 (fitted on 215 sites with real visit data, half of the sites within 15% error); 254 payments, revenue of $20,161 and 103 subscriptions come from TrustMRR. The public page does not give the shares of the second and third channels.

Supporting evidence: The three Show HNs on Hacker News all scored very low (11, 1 and 4 points); the Product Hunt launch got about 10 votes; the founder's X account has 709 followers; a third-party plugin marketplace has a skill that calls its API. The profile notes say organic search is about 52% and social about 10%, but the public page has neither number, so I used them only as a reference for the split and not as real figures. The profile also mentions that the 90% of visitors from Japan may come from a one-off repost or crawlers, and the reason was not found.

Inference: Splitting the two channel circles across four routes: direct 1,236 = subscribed customers 500 + forum launches 200 + searching the brand after open source tools 536; brand searches, social links and so on 2,764 = subscribed customers 100 + forum launches 600 + X posts 500 + searching the brand after open source tools 1,564. That 5% from forum launches, 5% from X posts and 30% searching the brand after hearing of it, and the way of splitting, are guesses. Visit counts are estimated, and both the +275.3% month-over-month and Japan at 90% are very unusual, so real visits may well differ; about 6.4 payments per 100 visits is only an order of magnitude.

Revenue, last month
$20,161
#91 of 563 by revenue
vs prior month
▲ 36%
$14,789 → $20,161, gain +$5,372
Website visits, last month
4,523
estimated from global rank
Similarweb global rank
4,529,488
Take-off year
2026
Team: small team
How it startedNot foundNo credible evidence of how it started
New-user source nowNew users mostly come from YouTube, X, Reddit, TikTok and the like
Growth engineNot foundNo credible growth evidence
Spreads purely by itselfMixed

Revenue over time

$8454–6 months ago avg
$1.4K3 months ago
$14.8K2 months ago
$20.2KLast month

Real payments synced by TrustMRR, in USD, by calendar month, renewals included.

MRR
$29,421
Revenue, last 30 days
$53,286
All-time revenue
$86,225
Active subscriptions
103
Payments via
Stripe
Source
TrustMRR, synced from the founder's payment account

1. Productwhat it sells, to whom, and how it charges

Product
Runs black-box penetration tests and red-team exercises on voice AI, AI agents, and websites using autonomous attack agents, billed per scan or by monthly subscription. It also offers its own uncensored attack/defense model Pingu Unchained, a desktop command-line tool AudnCode, and an open-source adversarial learning framework.
Users
Startups about to launch voice AI or AI agents, managed security service providers (MSSPs), and enterprise security teams
Positioning
Entered through voice AI red-teaming, using a specially trained attack model and a low-barrier $99 one-off scan, positioned against per-project manual penetration testing.
Pricing
One-off scan $99 (struck-through price $299, launch discount); WARLOCK $999/month; add-on packs Necromicon $179/month, GODZILLA $79/month; KONG $20/month; Pingu Unchained 10 $20/month; enterprise, MSSP, and API quotes Billing: mixed Source: Homepage
Prices on site
$99 per scan one-off (struck-through $299, launch discount); WARLOCK $999/mo; Necromicon $179/mo add-on; GODZILLA $79/mo add-on; KONG $20/mo; Pingu Unchained 10 $20/mo; Enterprise/MSSP/API quoted
Team
small team
Sharing mechanics
open source Partial: 'MetaClaw... an open framework for adversarial continual learning' and links to GitHub repos. Benchmark claims ('#1 @ Refusal-Benchmark') are not a user leaderboard.
“Partial: 'MetaClaw... an open framework for adversarial continual learning' and links to GitHub repos. Benchmark claims ('#1 @ Refusal-Benchmark') are not a user leaderboard.”Quote from the site

2. Channelswhere new users come from and what the traffic looks like

New-user source: Social media (Organic social). Similarweb last month: Direct 30.9%, Organic Search about 52%, social about 10% (X), with last month's visits up 275% month over month, but the country mix is 90% Japan and the reason was not found; the search terms are all brand-related (audn.ai download, audn.ai platform, audn.a harness, özgür özkan ai). Founder Ozgur Ozkan posted Show HNs on Hacker News (Pingu Unchained and Agentic Adversarial testing in 2025-11, Penclaw in 2026-02), launched on Product Hunt on 2026-03-12, and published scanner comparison posts on the blog (one was shared to HN by someone else in 2026-08); the TrustMRR description mentions a paid pilot with a YC S25 voice AI startup. The founder did not fill in marketing channels on TrustMRR. Existing users are retained through monthly subscriptions and add-on packs.

Last month's traffic sources

Search 52%Social 10%Direct 31%Unknown 7%

Top channel measured by Similarweb, others estimated by rank

Five-class attribution

Attributable new users, last monthShare
SEO10%
10%
Other79%

Direct + brand search 72%. Direct and brand search count as viral only when the site shows sharing mechanics; the largest share wins, under 5% means Other.

Countries

CountryShare of visits
Japan90.05%
Brazil9.95%

Language

English

Social sources

X

Keywords

17 keywords · brand ~72% · organic 100% / paid 0%

Similarweb top 5 keywords
audn.ai download
audn.ai platform
audn.a harness
audn.ai download for harness
özgür özkan ai

Non-brand queries: özgür özkan ai

Referrer categories

CategoryShare of referrals
Computers Electronics and Technology - Other12.63%

Site metrics (Similarweb, last month)

Visits MoM
+275.3%
Country rank
Japan #430053
Bounce rate
71.61%
Pages / visit
1.15
Top 3 channels
Direct › Organic Search › Organic Social (top channel 30.91%)

Growth-path traffic modelHow many people passed each step from launch to take-off

🔒 Unlock with quietmoney ProSee how many people passed each step from launch to take-off

3. Growth pathhow the first users arrived and what kept it growing

Classification

Basics

Timeline

Growth loop

Revenue and scale clues

Confidence and gaps

🔒 Unlock with quietmoney ProSee the timeline, growth loop, revenue clues and every source

Other sites that started via “Not found”

Similar sitessame category, also growing via Social media

FAQ

How much revenue does Audn make per month?

About $20,161 last month (real payments synced by TrustMRR), #91 of 563 TrustMRR sites; the month before was $14,789.

Where do Audn's new users come from?

New-user source: Social media (Organic social). New users mostly come from YouTube, X, Reddit, TikTok and the like Similarweb last month: Direct 30.9%, Organic Search about 52%, social about 10% (X), with last month's visits up 275% month over month, but the country mix is 90% Japan and the reason was not found; the search terms are all brand-related (audn.ai download, audn.ai platform, audn.a harness, özgür özkan ai). Founder Ozgur Ozkan posted Show HNs on Hacker News (Pingu Unchained and Agentic Adversarial testing in 2025-11, Penclaw in 2026-02), launched on Product Hunt on 2026-03-12, and published scanner comparison posts on the blog (one was shared to HN by someone else in 2026-08); the TrustMRR description mentions a paid pilot with a YC S25 voice AI startup. The founder did not fill in marketing channels on TrustMRR. Existing users are retained through monthly subscriptions and add-on packs.

How much does Audn cost?

One-off scan $99 (struck-through price $299, launch discount); WARLOCK $999/month; add-on packs Necromicon $179/month, GODZILLA $79/month; KONG $20/month; Pingu Unchained 10 $20/month; enterprise, MSSP, and API quotes

Dossier confidence: low. Parts 1 and 2 come from reading the site and Similarweb data; part 3 from public reporting, founder posts, Wayback snapshots, WHOIS and similar, with speculation marked.

The numbers: revenue comes from TrustMRR: real payments synced automatically from the founder's Stripe or other payment account, summed by calendar month, renewals included, in USD. Site visits come from the Similarweb public page; where it shows none, they are estimated from global rank or the TrustMRR payment count, as noted on the page.

Data through last month. Page updated 2026-10-31.