Terms of Service

Name

FRIENDIFY LTD

Registration

Company number 16656938

Registered office

128 City Road, London, United Kingdom, EC1V 2NX

ICO

ICO registration reference ZC025171

Contact

[email protected]

By accessing or using Friendify, starting a guest chat, creating an account, subscribing, connecting Discord, enabling voice, requesting AI-generated images, enabling character check-ins, sharing a journey card, sending a creator tip, listing or remixing marketplace content, or using privacy tools, you agree to these Terms. If you do not agree, do not use the service.

• Do not use Friendify for illegal activity, exploitation, harassment, hate, threats, self-harm encouragement, spam, malware, phishing, fraud, doxxing, impersonation, or privacy-invasive monitoring.
• Do not upload, request, sell, or distribute content that violates intellectual property, publicity, privacy, platform rules, or applicable law.
• Do not attempt to bypass usage limits, age gates, account controls, payment obligations, model safety systems, moderation, rate limits, or security measures.
• You must not bypass mature gates or misrepresent age to access mature or NSFW content.
• You must not upload adult, nude, or sexually explicit visual assets as SFW content or attempt to evade automatic or manual asset moderation.
• Do not paste secrets, provider keys, Discord bot tokens, card details, government IDs, or other sensitive data into fields not designed for that data.
• You must not request, generate, or store AI-generated images depicting minors or child sexual abuse material, non-consent, incest, bestiality, or the likeness of a real identifiable person, or otherwise use in-chat image generation to create illegal or exploitative content. Such requests are screened and may be refused before generation.
• Do not use personas, group chat, character check-ins, relationship cards, marketplace shares, referral links, notifications, reviews, reports, remixes, or creator messages to impersonate, deceive, harass, expose private information, or evade content and safety rules.

You must be able to form a binding contract and use Friendify only where the service is lawful for you. Mature marketplace content, NSFW chat, and adult image generation are strictly for users aged 18 or older. You are responsible for your account, OAuth provider security, email magic links, sessions, and activity under your account.

Mature marketplace content and similar experiences may require self-declaration or age assurance depending on the user’s apparent region and applicable law. Users who cannot complete a required mature gate must not access age-gated areas.

Friendify may offer a limited try-before-signup chat after the guest accepts these Terms and the Privacy Policy. Guest age is treated as unknown: the experience uses youth-safe settings and operator-selected, approved public SFW starter characters. A guest receives a pseudonymous profile and secure session and is subject to server-side message, conversation, maturity, model, and write limits. Guest access may be disabled by the operator at any time.

If you create or sign in to an account while the guest session is still available, Friendify may transfer the guest conversation and messages to that account and retire the guest profile. Abandoned guest data is deleted under the retention period in the Privacy Policy. Guest access does not unlock marketplace publishing, mature content, paid features, or registered-account actions.

Friendify provides a web-first AI character and roleplay platform with guest and account chat, single and group conversations, character creation and assistant suggestions, chat and image model selection, user personas, requested and optional spontaneous AI photos, memory, lorebook, Smart Chat Insights, relationship milestones and diary, optional character check-ins, limited journey share cards, collections, marketplace listings, remixes, creator profiles, notifications, reviews, reports, optional Discord, optional voice and real-time call mode, creator tips, privacy controls, and Stripe-hosted billing.

Feature availability, provider availability, usage limits, and plan entitlements may depend on configuration, safety controls, account status, subscriptions, and provider uptime.

Friendify output, including AI-generated text replies and any selfie-style character images you request in chat, is AI-generated and may be inaccurate, incomplete, unexpected, offensive, or inconsistent. Do not rely on Friendify for medical, legal, financial, emergency, safety-critical, or professional advice.

You are responsible for reviewing content you create, publish, share, list, or route through Discord. Friendify may moderate, remove, limit, or disable content and accounts to protect users, providers, the service, and legal obligations.

Friendify screens chat messages for a small set of serious safety signals — suicide and self-harm, eating-disorder encouragement, grooming and sexual content involving minors, and threats, bullying or hate. The screening runs on Friendify's own servers; message text is not sent to a third party for it.

Where screening is enforced, a message raising one of these signals may cause the character's reply to be replaced with an out-of-character safety message instead of roleplay. For suicide, self-harm or eating-disorder signals this points to UK support services. For grooming or sexual content involving a minor it stops that exchange. Receiving a safety message is not an accusation, is not a penalty, and does not by itself restrict or suspend your account.

Friendify is not an emergency service and its safety features are not a substitute for one. They can miss things and they can misfire. If you or someone else is in immediate danger, call 999. In the UK you can also contact Samaritans on 116 123 at any hour.

In a conversation you may request an AI-generated image of a character. Your optional instruction, the conversation context, and the character's stored appearance prompt are used to produce the image, which is routed to an image model through OpenRouter or, for adult images, through ModelsLab. Image generation is included on every plan and metered per plan, currently 3 images per 24 hours on Free and, per billing period, 30 on Spark, 150 on Bond, and 500 on Realm; the current allowance is always shown on the Pricing and Billing pages, which control if there is any conflict with these numbers, and a refused request refunds the quota it used.

If you enable the optional Spontaneous photos permission in Privacy settings, a character may also send you a photo without a button press: after a reply, an automated routing step reviews the recent conversation to decide whether a photo fits the moment and what it should show. Spontaneous photos use the same screening, moderation, storage, and per-plan image quota as requested images, never run once your image quota is spent, pause after a recent photo, and stop as soon as you turn the permission off. The permission is off by default.

Before generation, your instruction is screened by content moderation. Illegal requests, including child sexual abuse material, non-consent, incest, bestiality, and real-person likeness, are rejected. SFW characters require a clean result; NSFW images are produced only for NSFW characters accessed under a valid mature gate and are never produced for content classified as blocked. Every generated image is scanned before it is stored.

Generated images are stored to your account and linked to the conversation: SFW images in Friendify's standard blob store, and adult (NSFW) images in a private, EU-region storage bucket served only to you through an authenticated, mature-gated link. They are deleted when you delete the image, the conversation, or your account, subject to soft-delete recovery windows and any safety or legal hold. You are responsible for how you use images you generate, and Friendify may moderate, restrict, or remove generated images and the access behind them.

A user persona is private user-authored context that tells a character who it is speaking with. You are responsible for persona names, descriptions, and traits. Do not use personas to impersonate a real person, mislead other users, or provide instructions intended to bypass model or safety controls.

Smart Chat Insights may create summaries, mood and relationship labels, safety flags, and a short relationship diary. Milestones may be recorded from message-count thresholds, anniversaries, and forward relationship changes. These outputs are AI-assisted or automatically derived and may be inaccurate; they are product context, not factual, medical, psychological, or professional assessments.

Character-initiated messages require the proactiveMessages permission, operator enablement, and an eligible conversation. They use configured quiet periods and cooldowns, create an in-app notification, and can be muted globally or per conversation. Turning the permission off stops future check-ins but does not automatically delete messages already delivered.

A relationship share card is created only at your request and is limited to a small SFW allowlist of journey aggregates. Mature or safety-flagged conversations are not shareable. Anyone who receives a valid link may view the limited card, so you are responsible for choosing where to share it. Delete or restrict the underlying conversation if the link should no longer resolve.

You keep ownership of content you provide, subject to rights held by others. You give Friendify a worldwide, non-exclusive, royalty-free license to host, process, display, transmit, adapt, and use your content only as needed to operate, secure, improve, moderate, and support the service.

For marketplace content, creator profiles, listings, reviews, reports, and public character pages, you allow Friendify to display that content to other users according to your selected visibility settings and marketplace rules.

Creator profiles, character listings, collections, reviews, reports, and marketplace pages must be accurate, lawful, and safe for the selected visibility and maturity level.

Friendify may use automatic visual classification and manual review to classify uploaded character, cover, and gallery assets as mature, keep assets in review, reject assets, remove listings, restrict distribution, require self-declaration or stronger age assurance, disable mature content access, or suspend marketplace privileges when content, age status, rights, safety, or policy issues arise.

If an uploaded visual asset is classified or reviewed as mature, it may not be published as SFW marketplace content. You may need to mark the listing mature, replace the asset, or wait for review before public distribution.

Mature/NSFW access may be unavailable if Friendify cannot lawfully or reliably provide it in a region.

Installing a marketplace listing creates an account-owned copy. Remixing creates an editable derivative linked to the source listing so provenance and remix counts can remain visible. You must have the rights needed for any content you remix, and a remix does not transfer ownership of the source creator's content or brand.

Marketplace share links may include a validated Friendify user identifier for referral attribution. Friendify may record referral clicks, installs, follows, reviews, remixes, and listing milestones and use them for discovery shelves, badges, creator analytics, service notifications, and abuse prevention. Do not manufacture activity, referrals, reviews, follows, installs, tips, or rankings.

Where the Online Safety Act 2023 or similar safety law applies, Friendify uses proportionate account controls, rate limits, mature-access gates, prompt and generated-image screening, automated visual classification, manual review where appropriate, authenticated storage, user reports, and enforcement actions to reduce illegal content and misuse. Automated tools may refuse a request, classify an asset, restrict public display, or hold content for review; they do not guarantee that all harmful or illegal content will be detected.

Users and affected people can report a marketplace listing, review, creator surface, generated asset, or other content through the available in-product report control or by emailing [email protected] with the URL or content identifier and a concise explanation. Do not resend illegal material when a URL, identifier, or description is enough. Friendify may preserve evidence, restrict access, investigate, remove content, suspend accounts, or notify competent authorities where required.

If you believe Friendify applied a content, account, or safety decision incorrectly, email [email protected] with the decision, affected URL or account, and reasons for review. We will handle reports and complaints under applicable law and our documented safety procedures. Data protection complaints follow the separate process in the Privacy Policy.

Discord is optional. Your Friendify website account remains primary, Discord account linking is separate from bot installation and channel setup, and bot tokens or runtime secrets must stay server-side.

Voice is optional and consent-controlled. Friendify uses ElevenLabs for both speech-to-text and text-to-speech, including a real-time, turn-based voice call mode that captures short microphone audio, transcribes it, and streams character audio back. Your microphone audio and transcripts are processed by ElevenLabs in the United States; because zero-retention mode is not available on Friendify's current plan, ElevenLabs may retain that data under its own privacy policy, as described in Friendify's Privacy Policy. You are responsible for obtaining any consent required by local recording or voice laws before using voice features, real-time calls, or transcription with other people.

Friendify plans are Free, Spark, Bond, and Realm. Paid plan prices, taxes, renewal timing, entitlements, and limits are shown on the Pricing page or Stripe-hosted Checkout at the time of purchase. The Pricing page controls if there is a conflict with older copy.

Payments and subscription management use Stripe-hosted Checkout and the Stripe customer portal. Friendify does not collect card numbers, CVC/CVV, expiry, or raw payment method details.

Subscriptions renew unless cancelled before renewal through the customer portal or another supported flow. Access generally continues until the end of the paid period unless the account is terminated for breach, fraud, safety risk, chargeback, or legal reasons.

Refunds are handled case by case and according to applicable law, consumer rights, Stripe processing rules, and any specific refund notice shown at purchase. Statutory rights are not limited by these Terms.

One-time creator tips are separate from plan subscriptions and are covered by the Creator Tips section.

Creator tips are voluntary, one-time payments to support a creator through Stripe-hosted Checkout in payment mode, with an optional message of up to 280 characters. A tip is not a subscription and does not buy any product, service, entitlement, or additional access. Friendify does not collect card numbers, CVC/CVV, expiry, or raw payment method details.

Friendify is the merchant of record for tips. There is no Stripe Connect: Friendify receives each tip into its own Stripe account, retains a platform commission applied at the time of payment, and pays creators their share separately on a manual or batch basis. A tip is confirmed by Stripe's signature-verified checkout.session.completed webhook, and sending a tip does not create any contract between you and the creator.

Any message sent with a tip is subject to the Acceptable Use rules and may be moderated, limited, or removed. Do not use tips or messages for illegal payments, harassment, impersonation, or to send prohibited content.

Because a tip is a voluntary payment carried out immediately with your consent, completed tips are generally non-refundable, except where a refund is required by consumer law or where a payment failed or was made in error. Friendify may refuse, reverse, withhold, or refund a tip, adjust a creator payout, or suspend tipping in cases of chargeback, fraud, error, abuse, safety risk, or breach of these Terms. Creators are responsible for their own taxes on tips they receive, and nothing in this section limits your mandatory statutory rights.

The Privacy Policy and Cookie Policy explain how Friendify handles personal data, cookies, local storage, privacy requests, consent controls, international transfers, and processors.

You must not use Friendify to collect, infer, publish, or process personal data about others unless you have a lawful basis and all required notices, consents, and permissions.

Friendify may update, suspend, limit, or discontinue features for security, legal, provider, safety, pricing, operational, or product reasons. We try to provide notice for material changes where reasonable or legally required.

Provider outages, model changes, Discord API changes, Stripe issues, Yoti availability, or voice provider limits may affect the service. Friendify is not responsible for third-party services outside its control.

We may suspend or terminate access if you breach these Terms, create legal or safety risk, misuse AI or integrations, evade limits, fail payment, violate marketplace rules, or harm other users, providers, or the service.

You may stop using Friendify at any time and can request export or deletion through the Privacy center where available. Account deletion includes a 30-day cooling-off period before irreversible deletion steps continue.

Friendify is provided on an as-is and as-available basis to the fullest extent allowed by law. We do not promise uninterrupted service, perfect accuracy, specific AI output, specific marketplace performance, or compatibility with every third-party platform.

Nothing in these Terms excludes liability that cannot legally be excluded. For claims that can be limited, Friendify's aggregate liability is limited to the greater of the fees you paid to Friendify in the three months before the claim or GBP 100.

These Terms are governed by the laws of England and Wales. Courts of England and Wales have jurisdiction, except where consumer law gives you mandatory rights to bring claims in your local courts.

Legal, support, privacy, billing, and safety questions can be sent to [email protected]. Include your account email and enough detail for us to identify the issue, but do not send passwords, Discord bot tokens, provider keys, card data, or unnecessary sensitive content.