Privacy

Last updated 17 October 2026

Reading needs no account. Guest personas, saved readings, and saved-for-later selections stay in this browser. When you read while signed in, your playthroughs, saved-for-later selections, and reading preferences sync privately to your account so you can continue on another device. Bringing an existing guest library into an account is an explicit choice. Optional feedback and email requests work as described below.

What is stored on your device

These live in your browser's local storage. They are not cookies. None is sent as it is stored, but the reading events described under Measurement below carry the random id, and a few of them say what a setting here is — for example the site that sent you, the reading size you picked, that you confirmed your age, and the test mark if you set one. Guest personas, including avatars, and saved playthroughs are stored in IndexedDB. Account reading saves also keep a browser copy for reliable saving when the connection is interrupted. These copies are separated by account; signing out clears that account's reading and saved-story caches. If another tab creates an unfinished saved-story change during sign-out, that account copy is retained and the interface reports it so you can sign back in and check the change. Clearing site data removes browser copies, while successfully synced account saves remain available after signing in again.

To return you to the same discovery view, this tab's browser history keeps its feed selection, filters, search text, loaded amount and visible position. A short-lived session-storage record connects ordinary links between public pages. These navigation records do not contain follow lists, saved readings or persona fields. Following uses an opaque account key to check that a saved position belongs to the current account; a different account starts at the top of its own list. Navigation snapshots are not sent to our servers or AI models. Their retention is controlled by your browser, including its ability to restore tabs.

The creator studio also keeps an account key, draft identifier, chapter reference and selected workspace panel in the current tab's history entry. After checking your account, it can reopen the saved draft and return to that location. This record contains no manuscript text, private reference notes or sign-in credentials; it is not placed in a public link or sent to analytics or AI models. Device-copy recovery remains separate, and account changes do not reopen another owner's draft.

What you send us on purpose

Author world and character reference. You can save private world rules, character relationships, the reader’s authored role and continuity notes for each manuscript edition. These notes are separate from reader personas, public copy and the original starting plan. A new private update begins with the approved edition’s reference; editing it does not alter the earlier edition. The reference is not part of the reviewer or public story payload. AI next chapter receives a saved snapshot only when you select Include saved world & characters. We retain its revision with the private proposal so a later change cannot silently enter that request. Unsaved reference edits are kept in account-specific device recovery copies.

Creator chapter proposals. When you choose AI next chapter in the creator studio, we send your story setup, selected chapters and direction to apikey.fun. Your original starting reference is included only when you select it. Reader personas and reading history are excluded. The result stays in a private account proposal until you add it or discard it. Unfinished instructions and proposal edits are also kept in your browser’s account-specific device copies. Adding a proposal saves the chapter and records AI assistance with the manuscript. Checking or playing a proposal does not request more AI. Story publication still requires review. Minimal request receipts prevent duplicate chapters after interrupted saves; discarded proposal text is cleared from the account, while separately retained device copies remain until you dismiss them.

Account personas. When you save a persona while signed in, we store its name, description, optional pronouns, avatar, default preference, and story selections with your account so they are available when you sign in on another device. They are private to your account and are not published as a profile or credit name. Avatars are resized before upload and stored privately on Cloudflare. Guest personas are uploaded only when you choose to import them. The account-import panel lets you choose saved readings, persona names and avatars, private reference notes, and reading settings separately. Private notes start unchecked; you must select them to upload existing guest notes. Later note imports fill empty account fields and keep existing account notes. You can edit or delete personas in the Personas library. A saved playthrough retains its original role and avatar even if you later edit or delete that persona, unless you explicitly choose to clear its saved identity when deleting it. That privacy option removes the persona name, avatar and link from matching self-insert playthroughs in the current library, while preserving choices, progress and authored character roles. Minimal records of the removed persona IDs prevent stale saves from restoring those identities. Other devices apply account cleanup when they reconnect. Separate guest copies and older downloaded files are not erased.

Account reading saves. When signed in, we save each playthrough's story, role snapshot, chapter position, selected choices, and story variables with your account. We also save your chosen theme and text size. These records are private and are not a public profile. Browser copies let you keep reading if syncing is interrupted; the interface tells you when a save has reached your account. Different journeys and conflicting device saves are kept separately. Signing out returns to the separate guest library on this browser; it does not delete confirmed account saves. Importing a guest playthrough leaves its original in the guest library, which does not sync. If that guest copy changes, importing it again creates a separate account playthrough.

Account-library downloads. While signed in, you can download your confirmed account playthroughs, personas, saved avatars and reading settings as unencrypted JSON. You can exclude private notes and avatars. This is a portable library file, not a download of your credit name, feedback, community contributions, deletion requests or other site records. The export stops if your account changes, data cannot be retrieved or reading changes remain unsynced. It can be restored into a guest library and then explicitly imported into an account.

Deletion requests. Delete individual account personas in My personas; optionally clear their retained playthrough identities as described above. Signed-in readers can submit a deletion request in Account → Your data, choosing library records, personas and notes, credit-name and feedback records, the account, or published contribution credits. We save the request reference, selected scope, status and dates privately with the account so you and the NetFable operator can review it. A request is not automatic deletion. You can check its status and cancel while it is pending. Once review begins, contact us about changes to its scope. Completed means the operator has recorded the requested work as completed after review; the status control itself does not erase data. No automated email confirmation is sent. You can also contact us about records outside a library download. There is currently no automatic whole-account deletion button or published response deadline. Browser-only guest data and files you downloaded are separate copies; removing account records does not erase those copies from your devices.

Following stories and creators. When you follow a story or creator, we keep that selection and its date with your account so their published stories can appear in Following. Your follow list is private and is not shown to creators or other readers. Recent change receipts help safely retry interrupted saves. Following does not subscribe you to email or send a community message. Manage follows from the Following feed; they are separate from reading-library backups.

Account credit names. Your credit name is separate from private personas and playthrough names. The team sees names you include with feedback, and readers may see credit names on published contributions. Changing your account name leaves existing published story credits unchanged. We keep the current name version and up to 32 recent change receipts, containing request identifiers, request fingerprints, version numbers and dates, to recognize retries and conflicting saves. These receipts are private.

Private playthrough names. You can name a saved journey to tell your routes apart. Guest names stay in this browser; account names sync only within your account. These names do not change your story character and are not published, shown to creators, or sent to an AI provider. Library downloads include saved private names. Clearing a name removes that label, not the reading record.

Guest-library backups. My NetFable can download a JSON copy of the guest personas, avatars, reference notes, reading settings, and saved playthroughs in this browser. You can exclude reference notes (descriptions and pronouns), avatars, or both from the exported copy without changing the originals. Omitting avatars can reduce the file size. Exports and restores are limited to 25 MiB; an oversized export fails instead of creating an unrestorable file. These backup files are not encrypted. Backup preparation and restoration happen locally; the file is not uploaded to NetFable. If downloads are unavailable, you can explicitly copy the backup text instead. Keep these files and copied text private. A backup contains your saved choices and progress, not the story text for offline reading. Saved playthrough identities retain their own names and avatars, including when you import readings without adding their personas to the editable persona library. Restoring a backup adds data without replacing conflicting saved journeys; upload to an account still requires your explicit import choice.

You can ask your browser to protect local storage from automatic cleanup. The browser may decline, and even protected storage can be removed by clearing site data. A file backup or a confirmed account save provides a separate recovery option.

A few things on the site ask you for something, and all of them are optional. Tapping a chip — why you stopped reading, or whether a chapter was worth it — sends that one word, attached to the same random id everything else uses. The feedback box, and the request forms on the classics and feature-request pages, send whatever you typed, up to 600 characters, plus a name if you chose to add one.

Your own words at a choice, in the stories that offer it. Some stories let you type your own reply instead of picking an offered option. What you type, the options on screen and a few lines of the chapter around them are sent to our AI provider (apikey.fun, running GPT models) to interpret your words and write one short reply in the story's voice; the story then continues along a branch its author wrote. Nothing else about you goes with it — your persona, your saved name and your reading history never do — what you type is never included in our measurement, and it is never stored beyond answering you.

A story you create. If you use the Create page, the title, the text you write or paste, and any optional world/character starting plan go to our AI provider (the same one as above) when you choose Create AI draft. Write a chapter and Import a script do not request AI: your manuscript goes to your private NetFable draft, with your declared writing method. That declaration is reviewed with the story and is shown publicly as creator-provided information if published; it is not independently verified authorship. These direct-writing drafts require a public description and do not request an AI synopsis. When you send an AI-started draft or update for review without a public description, the provider may also receive its title and an excerpt from the edited script to draft a short description for the reviewer. A description you write is submitted with your draft and is not replaced by this suggestion. Account details, reading history and persona notes are not added to these requests. Preview and chapter navigation do not request AI. Your drafts are stored with your account and are visible only to you and, once you send one for review, to the person reviewing it — until it is approved and published, when it becomes a story on the site carrying your handle. What you paste is never included in our measurement. Published stories from the same account appear together on a public creator page. Each story keeps the credit name it had when submitted, even if you later use a different name. Two accounts may use the same name without sharing a creator page. We keep the link between your account and its public creator identity privately; the page does not expose your login identifier, email, drafts, personas or reading history. A creator page appears only after at least one story is published. You can request review of public contribution credits through Account → Your data.

Your creator profile. You can separately provide a public display name, bio, website or social link, and profile image. These fields are not copied from your login details or private reader personas, and they are not sent to an AI model. Saving keeps a private account draft. Sending for review makes that version visible to the reviewer. Only approved changes are included in a later website update; editing a draft does not change the published profile or existing story credits. A public profile still requires at least one published story. A link you provide is public after review; NetFable does not verify ownership of the linked site.

The browser crops your profile image to a square and removes file metadata before upload. We store that prepared image privately; approved images become public with the profile. Images used by a draft or retained approved version remain stored. Unused uploads are eligible for cleanup after a day. Private request identifiers, input hashes and saved revisions help confirm interrupted saves without duplicating them. Profile edits may also have separate recovery copies in this browser for the signed-in account. Other devices and downloaded text copies have their own lifecycles. Relevant profile records and public pages are included when we review account-data requests through Account → Your data.

AI draft requests keep an account-scoped request identifier, an input hash and a status so a connection failure can be retried without creating a second copy. A minimal removed status remains after deleting a private draft, so an old request cannot recreate it. These records contain no additional copy of your story text and are included when reviewing relevant account-data requests.

Your original scene and starting plan are saved privately with a successfully generated draft, so you can view or copy them later. This original reference is available through your account, including when you write an update. It is not published with the story, included in the reviewer preview, or automatically added to later synopsis requests. Ideas or details incorporated into the generated or edited manuscript can become public when that story is published. Editing the manuscript does not overwrite it. Deleting the original private draft removes its starting reference too; approved or retired story records follow the account-data review process. Planning text is not included in analytics, and personal reading personas are not copied into it.

Story covers you upload. The studio prepares a cropped still image and removes image metadata before saving it privately with your draft. The original file is not uploaded, and cover images are not sent to an AI model. Reviewers can see the saved cover with your submission. Approved covers become public when the story is published. Replacing or removing a cover in a private update does not change the published edition until that update is approved and released. We retain images referenced by submitted and approved versions as review history. Removed draft images and failed uploads are queued for cleanup; account requests and published contributions are reviewed through Account → Your data.

An email address, if you ask us to email you. At the end of a story that is still being written you can ask us to email you when the next chapter is out. If you do, we keep your email address, the story you asked about and when you asked. The address is encrypted where we store it, and it is used for that one purpose: nothing is sent to it until you confirm it from a link in the first message, an address that is never confirmed is deleted after a week, and after that we write to you only when a new chapter of that story is out. Every email carries a link that deletes your address for good, and it works without signing in. Or write to us at the address at the foot of this page and we will delete it.

A name you type there is self-declared and proves nothing; if you happen to be signed in, the note is linked to that account so we can reply. Nothing you write is published anywhere without asking you first. Write to us at the address at the foot of this page to have a note deleted.

What we do not collect

• No account is needed to read, and we do not store a password. Email addresses provided for sign-in or chapter updates are used for those requested services.
• No payment details — nothing on NetFable costs money.
• Persona names, descriptions, pronouns, and avatars are not included in reading analytics. Account persona storage is separate from your NetFable credit name.
• No advertising, and nothing sold or shared with data brokers.

Measurement

We look at anonymous, aggregate reading data — how many people start a chapter and how many finish it — to work out which stories are worth continuing. It is counted per random id that your browser keeps until you clear site data (the netfable.sid.v1 above), never per person. With it goes the kind of device you read on — phone, tablet or computer, judged from whether the screen is a touch screen and how big it is — never its make, its model or its exact size. If the page runs into an error in your browser, it sends us the error's message, where it happened and which screen you were on, at most three a visit, so we can fix it.

We also count library section openings, resumed chapters, backup export and restore actions, and whether a browser-storage protection request succeeds. These events do not include persona names, notes, avatars, private playthrough IDs, or backup contents.

We also record the host name of the site that linked you — reddit.com, not the page, the search terms or anything after the slash — so we can tell which communities send readers who stay. A link we post ourselves may carry a plain tag such as ?via=tiktok, which does the same job more accurately. Some sites label their own links, as ChatGPT does with ?utm_source=chatgpt.com, and send no referrer at all; we record that plain name in the same place.

Google Analytics tells us roughly where visitors arrive from and on what kind of device. It runs on every page and sets a cookie to tell one browser from another; we do not ask first, and there is nothing on the site to switch it off. If you would rather not be counted, your browser can refuse it — see the cookie policy for how.

Our host, Cloudflare, also measures page performance without cookies and without identifying anyone.

Where the site runs

NetFable is served by Cloudflare. Your request reaches their network, which means your IP address is processed to deliver the page, as with any website.

A note you send through the feedback box, or the request form on the classics or feature-request pages, is forwarded to our team chat (Lark) so that somebody reads it. What you wrote and the name you chose go with it, and nothing else does.

Email is sent through Resend (resend.com), a delivery service. When we send you something — a sign-in link, or the confirmation and chapter emails described above — the address and the message pass through them to reach you. That is the only reason either reaches them, and it is the only place your address goes besides our own database.

Your choices

• Block or clear the analytics cookie in your browser settings, or use any content blocker — see the cookie policy.
• Clear site data to erase device personas and reading progress. Delete account personas from the Personas library while signed in; clearing the browser does not delete them.
• Private windows may remove browser copies when the private session closes. Account saves that synced successfully and files you downloaded remain available separately.
• Unsubscribe from chapter emails with the link in any of them. It deletes your address and works without signing in.
• Sign out to drop the session cookie, or simply never sign in — reading works the same either way.

Questions about anything here: hello@netfable.com.