How We Verify Candy Shorts
2 separate anonymous sessions read every public Candy Shorts page and must agree before anything is published. One browses as an iPhone, one as a desktop browser, both from a US connection with an empty cookie jar. Last full check: Oct 19, 2026.
Every number on this site comes from one place: a snapshot of Candy's public Candy Shorts pages, taken the way a first-time visitor sees them. This page explains how that snapshot is made, what goes into it, and where its limits are. If you want to check us, everything we read is public on candy.ai.
The two-session check
- Start clean. No Candy account, no login, an empty cookie jar. We see what an anonymous visitor sees, nothing more.
- Two different visitors. One session identifies as Safari on an iPhone, the other as Chrome on a Windows desktop. Both connect from a US data-center IP in New York, so prices and access reflect a US visitor.
- Read everything public. Each session reads the Candy Shorts lobby, every page of the full library, each official category filter, Candy's Brand New page, every public series page and the subscriptions page.
- Compare. The 2 snapshots are compared field by field: titles, slugs, tags, the Spicy flag, categories, synopses, the token cost of every episode, its lock state and whether a free video is present. They must agree. If they do not, nothing is published until a fresh pair of runs does.
- Compile. The agreed snapshot is compiled into the catalogue the site reads. Totals, medians and the free rule are computed by code from it, never typed in.
What we read, field by field
| Candy page | Fields read | Used for |
|---|---|---|
| Library (candy.ai/candy-shorts/categories/all) | Series id, title, order, card tags, Spicy flag, poster, "Sign up to view" state | Series count, Candy's order, the Spicy shelf |
| Category filters | Which series sit in each official category | Tag pages and the tag table |
| Brand New page | Which series Candy marks as brand new | The catalogue log |
| Series page | Title, synopsis, cast ("Chat with the Cast") and, per episode: number, token cost, original token cost, lock state, poster, video address for free episodes, subtitle tracks | Spec sheets, the free rule, tokens to finish, discounts |
| Subscriptions page | Plan cards, billing lines, the Premium benefits list, the "Cancel any time." line | Price table, Premium tokens per month |
| Shorts paywall | The "With subscription you get access to" list | What Candy says Premium adds to Shorts |
How the free rule is derived
Each episode button on a series page tells a visitor three things: its token cost, its original token cost and a lock state. We sort every episode into one of three states with a fixed rule:
- Free, no account: no lock state and a playable video.
- Free with an account: zero tokens, but locked behind sign-up.
- Tokens: everything else, at the token cost Candy shows.
Counting those states across all public series gives the rule you see across the site:
| Episode | Access | Applies to | Note |
|---|---|---|---|
| Episode 1 | Free, no account | 148 of 148 | Plays here and on Candy |
| Episode 2 | Free with a free account | 148 of 148 | Sign-up only, no card |
| Episode 3 onward | 12 tokens each | 501 episodes | Most series |
| Episode 3 onward (premium series) | 30 tokens each | 14 series | All with a real-performer twin |
| Spicy series | Hidden until sign-up | 14 series | Blurred poster, title and tags only |
Discounts are caught the same way: when an episode's token cost is lower than its original cost, a discount is running. At the last check, no discount was active: token cost matched original cost on every episode.
Derived numbers
Tokens to finish is the sum of the token costs of a series' episodes. Median means the middle value when all series are lined up from lowest to highest, which is why the "typical series" (36 tokens) can differ from the average. Episode 1 runtime is read from the free episode 1 video file itself (its MP4 header), not taken from a label. The catalogue-wide figures (840 public episodes, 7,272 tokens for everything public, a median episode 1 of 1:06) all come out of the same compile.
How often we check
Weekly. Each run produces a new pair of snapshots; once they agree, the catalogue is recompiled, every page updates, and anything that changed since the previous check goes into the catalogue log with its date. The date of the last verified check is in the chip at the top of every page and in each byline: Oct 19, 2026.
Episode 1 frame checks
Counts do not tell you what is on screen, so for free episodes we also look ourselves. A frame check means we take 6 frames spread evenly from 10% to 85% of episode 1 and record a plain fact: whether there is nudity in those framen, with an optional one-line note and the date. Results appear on the series spec sheet under "Explicitness signals", marked "Our frame check".
15 of 148 public series have a published frame check so far. We only check free episodes. Paid episodes are never described on this site, because we would have to buy and watch them, and we do not publish what we have not seen.
What is not verifiable
Outside the method
- Hidden Spicy series. 14 series show anonymous visitors only a blurred poster, title and tags. Their episodes, prices and cast are not verified.
- Paid episodes. What they show is not verified and never described.
- Token pack prices. Candy's token page asks you to sign in.
- Premium and episodes. Whether a Premium member still pays tokens per episode or gets paid episodes included.
- Other visitors. Prices and access outside the US, for logged-in accounts or with personal offers may differ.
Policy facts and media
For questions like safety, billing and cancelling, we read Candy's terms of service, privacy notice, underage policy and help center, quote them word for word and link the document. Candy can change those texts; each of its legal pages shows its own revision date at the top.
The free episode 1 videos and posters shown here are Candy's own public files, the ones it serves to any visitor for free, stored on our server so they play without a Candy account. Nothing behind a paywall or a login is copied. Rights holders can reach us through the contact and DMCA page.
FAQ
How do you verify Candy Shorts data?
2 separate anonymous sessions, one as an iPhone and one as a desktop browser, both from a US connection with an empty cookie jar, read every public Candy Shorts page. The snapshots must agree field by field before the catalogue is published.
Why do you check Candy anonymously?
Because that is what a first-time visitor sees. It keeps our numbers reproducible: anyone can open the same public pages and compare. The trade-off is that anything behind sign-up, like hidden Spicy series, stays unverified.
How often is the Candy Shorts data updated?
Weekly. Last verified check: Oct 19, 2026. Every change between two checks is written to the catalogue log.
Why two sessions instead of one?
A single run can hit a glitch, an experiment or a page that loaded halfway. Two independent sessions with different browsers that agree on every field make that much less likely.
Do you watch paid Candy Shorts episodes?
No. We only watch and describe free episodes. We never describe what paid episodes show.
Why does Candy show me a different price?
Our prices are what an anonymous US visitor saw at the last check. Your country, your account, a discount or a personal offer can change what you see. Your own checkout page is what counts.