Welcome to VividSpark, operated by VIVIDTECH PTE. LTD., a company incorporated in Singapore ("VividSpark," "we," "our," or "us").
At VividSpark, we believe that creators should have confidence not only in the tools they use, but also in how their information is handled. This Privacy Policy explains what information we collect, how we use it, how we protect it, and the choices available to you when using www.vividspark.ai and our related products and services (collectively, the "Services").
By creating an account, accessing, or using our Services, you acknowledge that you have read and understood this Privacy Policy.
VividSpark is an AI-native creator growth platform that helps creators streamline their content publishing workflow through AI-powered music matching, music licensing, publishing, analytics, and related creative tools.
This Privacy Policy applies to:
It does not apply to third-party services, websites, or platforms that we do not own or control.
When you create an account, we collect:
Passwords are securely hashed before storage. We do not have access to your plaintext password.
To provide our Services, you may upload:
These files are processed solely for providing requested features such as: AI Soundtrack, background music matching, cover generation, publishing workflow, AI video production, and performance analysis. To provide these features, some of this content is sent to the AI service providers listed in Section 6.1.
Ownership of uploaded content remains with you.
Payments are processed securely by Stripe.
We do not collect or store full credit card numbers, security codes (CVV), or complete payment credentials.
We may receive limited payment-related information from Stripe, including payment status, subscription status, transaction identifiers, billing country, and partial payment method information (such as card brand and last four digits).
Stripe processes payment information in accordance with its own Privacy Policy.
We automatically collect certain information regarding how our Services are used, including browser type, device type, operating system, IP address, approximate geographic region, pages visited, session duration, feature usage, click interactions, referral sources, and error logs.
Certain AI-powered features require permission to analyze uploaded content.
When you use these features, we may process information extracted from uploaded videos, including scene segmentation, motion analysis, visual characteristics, timing information, music matching metadata, and AI-generated recommendations.
To produce this analysis, we may send the following to our AI service providers:
Which provider receives what, and for which feature, is listed in Section 6.1. This analysis is performed solely to deliver requested functionality.
If you authorize VividSpark to access supported social media accounts, we may collect performance information including views, engagement metrics, reach, audience interactions, and publishing performance.
This information is collected only after you grant the required permissions.
When you use the AI Soundtrack assistant, we keep a record of the conversation to provide the feature, investigate problems, and improve our products. This record includes:
These records are stored in our primary database. Access is limited to authorized VividSpark staff who need it to operate and support the Services. To generate replies and search the music catalog, your messages, or text derived from them, are processed by our AI service providers: Anthropic, to understand your messages and write replies, and OpenAI, to interpret music search requests (see Section 6.1).
The text of your messages and of the replies is deleted after 180 days. We keep non-content information, such as timestamps and which part of the feature responded, for longer to measure and improve the Services. Text you can see as the title of a saved soundtrack version stays in your version history and is deleted at your request (see Section 11). Your browser also keeps a copy of the conversation on your device for up to 24 hours.
When you use the Growth Agent's Leads feature, we collect and organize information about people and businesses who may be interested in what you offer (“leads”). Leads come from three places:
For each lead we store a short excerpt of the post or message (up to 300 characters), the name or handle the author shows publicly, a link to the original, when it was posted if known, and our assessment of it: what the person appears to need, a score, a one-sentence reason and, for lead search, whether their location appears to match the area your campaign serves. Notes you add are stored with the lead.
To make that assessment, the text of the posts and messages, together with your campaign's description, is processed by our AI service provider, Anthropic. We do not contact anyone on your behalf: whether and how to reach out to a lead is your decision.
Leads are visible only to you. A lead nobody has contacted is deleted automatically 90 days after it was last updated; a lead you have marked as contacted, replied or won is kept with your account. To avoid reading or paying for the same post twice, we keep a list of the items already assessed for each campaign; it contains identifiers only (such as a link or a fingerprint of the text), not the text itself, and is deleted after 90 days. Records of notifications received from our social media provider contain only an event identifier and type, and are deleted after 30 days.
If you believe a post of yours appears as a lead in someone's VividSpark account and you want it removed, contact us (see Section 17) and we will delete it.
Where this feature is available to your account, you can ask VividSpark to produce a finished video for you — editing your footage, adding music from our catalog, sound effects, on-screen text and, if you choose, an AI voice-over. To do this we process:
The work is carried out by an AI agent running in an isolated computing environment provided by Modal. The agent uses Anthropic's AI models, which receive your request, your answers, still frames and transcripts of your footage, and the brief. Voice-overs are generated by OpenAI, which receives the approved script. Your footage is also analyzed as described in Section 3.5.
The working files for a video, including the agent's session, are kept for about 24 hours so that you can ask for changes, and are then deleted. The text of the conversation is kept as described in Section 3.7. Finished videos are kept so that you can download them. Purchasing a music licence always requires your confirmation; the agent cannot buy one for you.
Some of what our Services produce is generated or substantially edited by AI, including cover images, captions and titles, music matching decisions, voice-overs and produced videos.
Where the law or a publishing platform requires it, we may add a visible label to AI-generated content, such as a notice at the start of a video or in a voice-over, and we may record in the file's metadata that it was AI-generated, together with our name and a content identifier. Platforms such as YouTube, TikTok, Facebook and Instagram ask creators to disclose realistic AI-generated or AI-altered content when they post it; when you publish content made with VividSpark, you are responsible for making the disclosures those platforms and the law require. We do not offer voice cloning.
Master is an option in AI Soundtrack in which an AI music director watches your video, chooses music from our catalog and edits it to your video. When you choose Master for a video, we process:
Anthropic's AI models receive the still frames, the transcript and the derived information, and use them to describe the video and to choose and edit the music. The audio itself is not sent to Anthropic; it is transcribed by OpenAI (or, if that fails, by Replicate). The music searches that the AI director writes, which can describe what happens in your video, are processed by OpenAI to search our catalog. Modal extracts the frames and audio, detects speech and renders the result. The name of your video file is not sent to these providers. See Section 6.1.
The still frames and extracted audio are stored as private temporary files and are made available to the relevant providers only through access-controlled, time-limited links. We normally delete these temporary files within 48 hours after they are created. We keep the transcript and a record of the steps Master took — including working notes and searches that may describe your video — for up to 180 days after the relevant session activity, unless a shorter period is requested or a longer period is required by law. We then delete the text or irreversibly de-identify it and retain only aggregate operational figures. The finished soundtrack is kept with your AI Soundtrack session until you delete the session or your account, subject to our backup-deletion schedule. Provider-specific retention periods and processing locations are listed in our AI Service Provider List.
We use collected information to:
We do not sell your personal information.
Depending on your jurisdiction, we process information based on one or more of the following legal grounds:
Where consent is required, you may withdraw it at any time, although doing so may affect certain features of the Services.
To operate VividSpark efficiently, we rely on carefully selected third-party service providers. These providers process information only as necessary to perform services on our behalf.
These providers run the AI models and processing behind our features. Each receives only what the feature you are using needs. For each one we list the company, what it is used for and what it receives, where it mainly processes data, and how long it keeps it.
/v1/audio/transcriptions) is not kept in OpenAI's abuse-monitoring logs; music search text sent for search (/v1/embeddings), and images and text sent to its other endpoints, may be kept in abuse-monitoring logs for up to 30 days, or longer where the law requires. OpenAI does not use API data to train its models by default.These providers may process data in the United States and in other countries (see Section 9). The retention periods above are those each provider publishes for the services and account settings we use, as of October 2026. We review this list against our contracts and account settings and update it, and notify you of material changes, when they change.
We use cookies and similar technologies to operate our Services, maintain user sessions, improve functionality, measure performance, and understand how visitors interact with our platform.
These cookies cannot be disabled without affecting core functionality.
We use Google Analytics (_ga, _gid, _ga_*) and Microsoft Clarity cookies to understand website traffic, feature popularity, and user experience.
We also set our own cookie, vs_ft, when you first visit our site. It records how you found us: the campaign tags in the link you followed (such as utm_source), the website that referred you, the page you first landed on, and, if you came from an ad, which ad platform it was. For Google Ads it also keeps the click ID Google adds to the link; for other ad platforms it keeps only the platform's name. We use it only to record, when you create an account, where that account came from. It expires after 90 days and is not shared with third parties.
Meta (Facebook) Pixel sets the _fbp cookie to measure marketing effectiveness.
The Reddit Pixel sets first-party cookies on our site to measure how ads shown on Reddit perform. At present these are _rdt_uuid, a randomly generated identifier for your browser, and, only when you arrive by following a Reddit ad link, _rdt_cid, which stores that ad's click identifier. Reddit's pixel decides the names of these cookies and how long they are kept, and may change them. The pixel reports the pages you visit on our site and, once, that you signed up — at the same moment as described for OpenAI below — with an internal identifier for your VividSpark account (not your email address). We do not send Reddit your email address, name or phone number.
The OpenAI Ads Measurement Pixel sets two cookies to measure how ads shown in ChatGPT perform: __oppref, which stores the identifier carried by a ChatGPT ad link (30 days), and __obref, a randomly generated identifier for your browser (365 days). When you enter contact details into a form on our site, such as your email address or name, the pixel converts them into a SHA-256 hash in your browser and sends only that hash with these measurement events, to help us measure ad performance. The details themselves are not sent as plain text.
When you sign up after following a ChatGPT ad link, our servers also report the sign-up to OpenAI through its Conversions API, so the ad can be credited even if your browser leaves the page before the pixel finishes. A sign-up is reported once, when it counts: for an email sign-up, when you verify your email address by opening the link we send you (an address that is never verified is never reported); for Google or Apple sign-in, when your account is first created. Signing in again later is not reported. That report contains an internal identifier for your VividSpark account (not your email address), the time the sign-up counted, the address of the page you signed up on (without its query string), the ad identifier from your __oppref cookie, the browser reference from your __obref cookie, and your browser's user agent. It does not contain your email address, name, phone number or IP address. We send it only when the browser that completes the sign-up (for an email sign-up, the browser in which you open the verification link) carries an ad identifier from a ChatGPT ad, and we do not store that identifier ourselves.
VividSpark does not currently provide a dedicated cookie consent interface.
Most browsers allow you to manage or disable cookies through browser settings.
Disabling certain cookies may affect the availability or functionality of portions of the Services.
Because VividSpark serves users worldwide, your information may be processed outside your country of residence. When transferring personal information internationally, we take reasonable measures to ensure appropriate safeguards are implemented, including contractual, technical, and organizational protections where applicable.
VividSpark currently does not provide a self-service account deletion feature.
To permanently delete your account, please contact us at support@vividspark.aior use the "Talk to Human" support option within the platform.
Upon receiving a valid request, we will delete or anonymize your personal information where reasonably practicable, except where retention is required by applicable law.
Depending on your jurisdiction, you may have the right to:
Although VividSpark has not yet implemented formal GDPR or CCPA compliance workflows, we respect legitimate privacy requests and will make reasonable efforts to respond in accordance with applicable law.
Requests may be submitted to: support@vividspark.ai
We implement reasonable administrative, technical, and organizational safeguards including encrypted password storage, HTTPS encryption, secure authentication, access controls, infrastructure monitoring, and trusted cloud infrastructure providers.
No system connected to the Internet can guarantee absolute security.
Our Services are not specifically directed toward children, and we do not knowingly collect personal information from children in violation of applicable law. Parents or guardians who believe a child has provided personal information may contact us at support@vividspark.ai.
Our Services may contain links to third-party websites or social media platforms.
We are not responsible for the privacy practices or content of third-party websites.
We may update this Privacy Policy from time to time. When material changes are made, we will provide appropriate notice through our website, email, or within the Services. Continued use of the Services after an updated Privacy Policy becomes effective constitutes acceptance of the revised policy.