- We never sell your data, and we don’t share it with advertisers or data brokers.
- Each provider receives only what its job needs, for example the prompt and reference image for one image, not your whole account.
- Providers process data under their data-processing and API terms. Where a provider lets us switch off training on your content, we do. Where it doesn’t (Runway’s terms, for example, say inputs and outputs may be used to train its models), its own terms govern that use and we have no control over it. The training columns below show each provider’s stance, and the Privacy Policy explains more.
- Every tool shows its price before you run it, and tools that let you choose a model show which one will run.
- Public timestamp services (DigiCert, and Sectigo for consent evidence packs) receive only a one-way fingerprint of a file’s Content Credential or of an evidence pack, to stamp the time: no content and no personal data.
- Our servers also download a few public files, and nothing about you goes with those requests: exchange rates (the European Central Bank’s, through the Frankfurter service), the public C2PA trust list (from GitHub), the invisible watermark’s model files (from Adobe), font files (from Google Fonts, asked for by font name only) and Runway’s published price list.
- Business customers: this list is Annex III of our Data Processing Agreement.
| Provider | What it does for us | Data it receives | Where |
|---|---|---|---|
| Supabase (Supabase, Inc.) | Database, sign-in, private file storage, the public pictures and video of published property pages, and live collaboration. | Account details, projects, uploads, brand assets, usage and purchase records, and everything else stored in your account. | Amazon Web Services cloud, in our project’s region |
| Web hosting provider | Runs the QikTools website and servers. | Every request to our site: IP address, browser details and server logs. | To be named here before QikTools goes live |
| Cloudflare (Cloudflare, Inc.) | Turnstile bot protection when you sign up, claim your trial credits, send a report, or send an enquiry through a property page. | IP address and browser signals used to tell people from bots. | Global network |
| GoDaddy (GoDaddy Operating Company, LLC) | Business email: sends and stores account, team, reminder and support emails. | Your name, email address and the message. | United States |
For US dollars, our payment partner acts as merchant of record: it sells to you as our reseller, so it is responsible for your payment data under its own privacy policy. For Indian rupees, we are the seller and Razorpay is our payment gateway: it collects your payment details itself and also handles them under its own privacy policy. We receive a record of what you bought, never your full card number.
| Provider | What it does | Data it receives | Where |
|---|---|---|---|
| Dodo Payments (merchant of record) | Sells plans, team seats and credits priced in US dollars to you as our reseller. Processes those payments, handles sales tax, VAT and GST, issues receipts and carries out refunds. | Your name, email address, billing address and country, payment method (card details go to it, never to us) and what you bought. | Its group companies (including in the United States and India) and its payment processors worldwide |
| Razorpay (Razorpay Software Private Limited) | Processes payments in Indian rupees for the plans, team seats and credits QikTools sells to you directly — UPI, cards, net banking and wallets, and UPI AutoPay or card and bank mandates for plans that renew — and carries out refunds. It is a payment gateway, not a reseller: QikTools is the seller of a rupee purchase. | Your name and email address, a phone number if you give one in its payment window, your payment method (card, UPI and bank details go to Razorpay, never to us), the amount and what you bought, and device and network signals it uses for security and fraud checks. | India (Razorpay, and the banks, card networks and UPI apps that process the payment) |
| Stripe | Card payments for team plans, where they are enabled. | Name, email address, billing details and payment method (card details go to Stripe, never to us). | United States and other Stripe locations |
These providers receive your prompts and uploads when you run a tool that uses them. Some keep request data for a limited period (for some, up to 30 days) to detect abuse, under their own terms.
| Provider | What it does for us | Data it receives | Where | Training on content |
|---|---|---|---|---|
| OpenAI (OpenAI, L.L.C.) | Writing, scripts, translation, plans, captions and document answers; reading images, pages and documents (for example product photos and alt text); GPT Image generation and editing; text-only transcription; the help assistant (in text and, with Talk to Munns, by voice); safety checks on requests; translating non-English ad scripts for the claim check; the check that decides whether an output needs a visible AI label; descriptions and search for your asset library. | The text, images or audio each job needs (for Talk to Munns, the live audio of the conversation). | United States | No training by defaultOpenAI’s Services Agreement (v. 1 January 2026): “OpenAI will not use Customer Content to develop or improve the Services, unless Customer explicitly agrees to such use.” |
| Fal (Features & Labels, Inc., fal.ai) | Runs image, video, voice, timed-transcription, voice-separation, music, lip-sync, 3D, background-removal and upscaling models, joins video pieces with their soundtrack, trains custom brand models, and hosts generated and working files. | The prompt and the images, audio or video each job needs; training images for a custom model; generated files. | United States | No training by defaultfal’s API Services Terms: “Company will not use Client Content to create, train, develop (directly or indirectly) Company’s products or services” — for models fal marks “Pending Enterprise Ready” its security commitments and data processing addendum don’t apply. We only use models fal marks Enterprise Ready; cloud film renders are prepared and checked on our own servers and joined with fal’s Enterprise Ready FFmpeg compose. |
| Runway (Runway AI, Inc.) | Video models, talking-head presenters, image models, the Voice Studio’s Expressive voice, voice changing, dubbing and voice isolation. | The prompt, the text to be spoken, and the images or audio each job needs. Presenter photos are deleted after each video. (Runway’s terms still let it use inputs, including presenter photos, to train its models.) | United States | May train on contentRunway’s Terms of Use (updated 15 September 2026, read again on 10 October 2026) say “Inputs and Outputs may be used by the Company to train and improve its AI models” — this covers everything we send through Runway, including other companies’ models it runs for us and the Nano Banana pictures it makes as a backup. A request made with “Only providers that don’t train on my content” never goes to Runway. |
Many models are made by other companies and run for us by Fal or Runway. Depending on the model, the content of a job is either processed by Fal or Runway, or passed on by them to the model’s developer to run it. These are the developers whose models may receive your content. The Where column shows each one’s location: Kling, ByteDance and MiniMax reach us through Singapore companies of groups based in China, and Alibaba Cloud is based in China.
| Developer | Models | Reached through | What they may receive | Where | Training on content |
|---|---|---|---|---|---|
| Gemini Omni Flash 1.1, Veo 3.1, Veo 3.1 Fast, Veo 3.1 Lite, Nano Banana 2.1, Nano Banana Pro, Lyria 2, Nano Banana 2 | Fal: Gemini Omni Flash 1.1, Veo 3.1, Veo 3.1 Fast, Veo 3.1 Lite, Nano Banana 2.1, Nano Banana Pro, Lyria 2; Runway: Nano Banana 2, Nano Banana Pro | Video, image and music jobs (Runway receives picture jobs only as a backup when Fal can’t make them, never for requests that ask for providers that don’t train on your content) | United States (Google may also process data in other countries where it has facilities) | No training by defaultGoogle’s Gemini API Additional Terms (effective 23 March 2026, page updated 28 April 2026, read again on 10 October 2026) say that for paid services “Google doesn’t use your prompts … or responses to improve our products”; we reach Google’s models through fal, which pays for them (fal marks the Nano Banana models we use “Enterprise Ready”, checked 10 October 2026), and, for Nano Banana pictures only as an automatic backup, through Runway — whose own terms still let Runway train on what passes through it. A request made with “Only providers that don’t train on my content” is never moved to Runway. | |
| ByteDance | Seedance 2.5, Seedance 2.0 (480p, 720p), Seedance 2.0 (1080p), Seedance 2.0 Fast, Seedance 2.0 Mini, Seedream 5 Pro | Runway: Seedance 2.5, Seedance 2.0 (1080p), Seedream 5 Pro; Fal: Seedance 2.0 (480p, 720p), Seedance 2.0 Fast, Seedance 2.0 Mini | Video and image jobs | Singapore (BytePlus Pte. Ltd., part of ByteDance, China) | Not statedByteDance’s BytePlus terms for these models are only readable when signed in, so we can’t confirm them; through Runway, Runway says it negotiates no-training terms with every model maker it runs (Runway’s own entry still applies). |
| Kuaishou | Kling 3.0, Kling O3, Kling O1 | Fal | Video jobs | Singapore (Kling AI Pte. Ltd., part of Kuaishou, China) | May train on contentKling AI’s Terms of Service (21 April 2026) let it use content to “create, test, improve, train” its models; we reach Kling through fal’s partnership, whose terms with Kling aren’t public. |
| MiniMax | MiniMax H3, MiniMax H3 Max, MiniMax Music | Fal | Video and music jobs | Singapore (Nanonoble Pte. Ltd., which runs MiniMax’s international services; MiniMax is based in China) | Not statedMiniMax’s API platform terms only load in a browser script and we couldn’t read them; we reach MiniMax through fal’s partner API. |
| Alibaba | Wan 3 | Fal | Video jobs | China (Alibaba Cloud); the region used for our jobs isn’t stated | No training by defaultAlibaba Cloud Model Studio (page updated 28 September 2026): it “will never use your data for model training”; we reach Wan 3 through fal’s partner API. |
| Black Forest Labs | FLUX.1 Kontext [pro], FLUX 3 | Fal | Photos to edit and editing prompts, and the pictures and prompts of video clips | Germany (BFL GmbH, Freiburg) | May train on contentBlack Forest Labs’ FLUX API Service Terms (4 August 2026, read again on 14 October 2026): “the Company may use Inputs and Outputs to train and improve its artificial intelligence models”; we reach FLUX.1 Kontext [pro] and FLUX 3 through fal’s partner API. |
| ElevenLabs | Eleven v3, Eleven v4, Sound Effects v2, Eleven Multilingual v2 (voice changer), Dubbing, Voice Isolator, Eleven v4 Turbo, Eleven Turbo v2.5, Scribe v2 | Fal: Eleven v3, Eleven v4, Sound Effects v2, Eleven v4 Turbo, Eleven Turbo v2.5, Scribe v2; Runway: Eleven v4, Eleven Multilingual v2 (voice changer), Dubbing, Voice Isolator | Text, voice recordings and audio | United States (Eleven Labs Inc.) | May train on contentElevenLabs’ Terms of Service (31 March 2026) let it use customer content “to improve the Services” unless the account holder opts out; the account holders for our requests are fal and Runway, not us, so the opt-out isn’t ours to set. |
| Sync Labs (sync.) | sync. lip-sync | Fal | Video and audio of the speaker | United States | May train on contentSync Labs’ Terms of Service (revised 6 June 2024) give it a licence to use uploaded content “to develop and improve the Service”; we reach it through fal’s partner API. |
| VEED | VEED Lipsync | Fal | Video of the speaker and the translated audio | United Kingdom (Veed Limited, London) | No training by defaultVEED’s API FAQ (read 7 October 2026): “We don’t use your API inputs or outputs to train our models unless you agree to it” (it keeps each request’s log for 7 days); its Terms of Use (22 April 2024) only let it train on content from free VEED accounts. Veed Limited is in London, United Kingdom; fal marks VEED Lipsync as a partner API, so we reach it through fal. |
| Creatify | Creatify Lipsync | Fal | Video of the speaker and the translated audio, only if Fal passes them on (Fal runs this model on its own servers) | United States (Creatify Lab Inc., California) | May train on contentCreatify’s Terms of Service (effective 23 September 2024) give it a perpetual licence to use what is sent and made “to provide, maintain, promote, and improve the Services”. fal’s own model page marks Creatify Lipsync as run by fal itself, not as a partner API, so your video and audio should stay with fal — which doesn’t train on them — and not reach Creatify; we list Creatify’s terms because they govern the model. Creatify Lab Inc. is in California, United States. |
| CassetteAI | Music Generator and Sound Effects | Fal | Music and sound prompts | United States (Pixl Technologies, Inc.) | Not statedCassetteAI’s terms (13 February 2023) don’t say whether prompts are used for training; we reach it through fal’s partner API. |
| Ideogram | Ideogram 3.0 | Fal | Image prompts | Canada (Toronto) | No training by defaultIdeogram’s Developer API Agreement: it “shall not use any User Input or User Output to train the Ideogram AI Model”, except content flagged as breaking its usage policies; we reach it through fal’s partner API. |
| Recraft | Recraft V3 | Fal | Image prompts | United States (Recraft Inc.) | No training by defaultRecraft’s Developer Terms (11 August 2026): “Recraft will not use any Assets created using the API Services to train Recraft AI models”; we reach it through fal’s partner API. |
| Tripo AI (VAST) | Tripo H3.1 | Fal | Product photos | Not confirmed (its site couldn’t be checked) | Not statedTripo’s site blocks automated reading, so we couldn’t open its terms; search listings of its help pages say paid users’ content isn’t used for training, but we haven’t confirmed that ourselves. We reach Tripo through fal’s partner API. |
| Hyper3D (Deemos) | Rodin Gen-2.5 | Fal | Product photos | United States (Deemos Corporation) | May train on contentDeemos’ terms say it may use prompts and outputs “to provide, maintain and improve our services, and conduct and support research”; we reach Rodin through fal’s partner API. |
| Bria AI | RMBG 2.0 | Fal | Photos for background removal | Israel (Tel Aviv) | No training by defaultBria’s background-removal terms (September 2026) let it use content only “for the performance and enhancement of the Products for Customer’s use only” and say it “shall not be used to train generative models” without written consent; Bria is based in Tel Aviv, Israel, its terms incorporate its data processing addendum, and fal marks this model Enterprise Ready, so fal’s own addendum covers it. |
| xAI | xAI Text to Speech | Fal | Text to be spoken | United States | No training by defaultxAI’s API security FAQ: it “never trains on your API inputs or outputs without your explicit permission”; we reach it through fal’s partner API. |
Models that Fal runs on its own systems, such as FLUX.1 [schnell], SeedVR2, Stable Audio 2.5, Demucs, DeepFilterNet, Chatterbox (by Resemble AI), TRELLIS, Qwen-Image and Z-Image, don’t send your content to their original authors. The full list of models is on the AI Models & Providers page.
- Google and jsDelivr: the Video Clipper, the AI Video Translator (to find where lip-sync applies) and the Real Estate Listing Studio (to blur faces) download an on-device face-detection model and its code. They see your IP address, not your video or photos.
- unpkg (backup only): the Reel Maker’s in-browser video-joining engine comes from our own servers. Only if that copy can’t be loaded does your browser fetch the same engine from unpkg, and it checks the file against a fixed checksum before using it. See Open-source software.
- Google and Dropbox (file pickers): when you choose a video from Google Drive or Dropbox in Video Dubbing Autopilot, your browser loads Google’s sign-in and file-picker scripts, or Dropbox’s Chooser. Only the file you pick is read.
- Cloudflare: the Turnstile check on the sign-up, trial, report and property-enquiry forms.
- OpenAI (Talk to Munns): when you talk to Munns out loud, your browser streams your voice straight to OpenAI over an encrypted connection that our server starts, so OpenAI sees your IP address as well as your voice. Your voice does not pass through our servers.
- Google (3D viewer): the 3D Product Studio’s viewer loads decoders from Google when a model needs them, and on Android “View in your room” opens Google’s Scene Viewer, which downloads the 3D file.
- Google and Microsoft (sign-in): when you choose “Continue with Google” or “Continue with Microsoft”, that company confirms who you are and shares your name and email address with us.
- Social platforms: when you open Instagram, TikTok, YouTube, X, LinkedIn or Facebook from a reminder to post, you go to that platform, which handles your data under its own policy.
Some tools can work with an account you hold at another company, but only after you grant access on that company’s own screen, and you can take it back there or in the tool. These companies act for themselves under their own terms and privacy policies. They aren’t our processors, and the access you give is not a sharing we decide.
- YouTube (Google): Video Dubbing Autopilot, once channel connections are switched on, reads your channel’s name, picture, uploads list and the caption tracks you choose, and adds a subtitle track only when you approve that upload. It never downloads your videos. See Video Dubbing Autopilot and connected channels.
- Meta, TikTok and Google Ads: the Winning Ads Loop, once ad-account connections are switched on, reads the ad-level results you ask for. It never creates, changes or publishes anything in your ad accounts. See Winning Ads Loop: ad results and ad-account connections.
- Google Drive and Dropbox: no standing connection. A file picker returns the one video you choose (see above).
We use providers’ data-processing and API terms, which require them to keep data secure and use it to provide their service. Where the GDPR or UK GDPR applies to a transfer outside Europe, we rely on the European Commission’s Standard Contractual Clauses and the UK Addendum. More in the Privacy Policy and the Data Processing Agreement.
We publish a change here, and in the feed below, at least 14 days before a new provider starts processing personal data for business customers, except when an urgent change is needed to keep the Service running or secure, when we publish it as soon as we can. Questions: hello@qiktools.com.
Get notified of changes for free
Add the feed to any feed reader or website-monitoring tool.
Payments in Indian rupees through Razorpay
Purchases priced in Indian rupees are now sold by QikTools itself and paid through Razorpay (Razorpay Software Private Limited, Bengaluru, India), a payment gateway: it receives the buyer’s name, email address, payment method (card, UPI and bank details go to Razorpay, never to us), the amount and what was bought, and device signals for its fraud checks. Plans that renew in rupees use UPI AutoPay or a card or bank mandate held by Razorpay and the buyer’s bank. Purchases in US dollars are unchanged: Dodo Payments sells them as merchant of record.
AI Video Translator: a new voice pipeline and lip-sync by VEED and Creatify
The AI Video Translator no longer sends videos to ElevenLabs’ dubbing through Runway. Each video’s speech is now separated from its background by Demucs (an open model Fal runs itself) and transcribed by ElevenLabs’ Scribe v2 (both through Fal), translated by OpenAI’s GPT-4.1 mini, and spoken by xAI Text to Speech — or ElevenLabs’ Eleven v4 Turbo for the languages xAI doesn’t speak, or Chatterbox in the speaker’s own or a Voice Vault voice — all through Fal. Lip-sync now uses VEED Lipsync for Standard (Veed Limited, London, United Kingdom, a Fal partner API, so the speaker’s video and the translated audio go on to VEED) and Creatify Lipsync for Premium and for the Standard lip-sync of short videos (Creatify Lab Inc., California, United States; Fal marks it as run on its own servers, so the video and audio should stay with Fal, and we list Creatify because its terms govern the model); Sync Labs’ lipsync-2 stays for the Close-up quality. VEED and Creatify are new model developers in the list below.
Transcription moves off OpenAI’s Whisper; new voice models
We no longer use OpenAI’s Whisper. The AI Video Clipper and its auto-captions transcribe with ElevenLabs’ Scribe v2 through Fal. Meeting minutes, podcast uploads and Video to Blog uploads now offer two transcripts: Quick (text only), made by OpenAI’s GPT-Transcribe with Scribe v2 as the backup, and Detailed (speakers and timestamps), made by Scribe v2 through Fal. Long recordings are sent to Scribe v2 in parts. The Natural voice now uses ElevenLabs’ Eleven v4 Turbo through Fal, with Eleven Turbo v2.5 as the backup. The Voice Studio’s Expressive voice now uses ElevenLabs’ Eleven v4 through Runway, which therefore also receives the text of those voiceovers; the same Eleven v4 through Fal is its backup. No new processor is involved.
Standard voices by xAI through Fal, and transcription changes
OpenAI is retiring its older speech models, so every stock voice — the Voice Studio’s Standard voices, voiceovers in the video studios, the Campaign Orchestrator and the API — now uses xAI Text to Speech, run through Fal; it receives the text to be spoken. Brand Guardian transcribes a video’s speech with ElevenLabs’ Scribe v2 through Fal. Text-only transcripts (Video to Blog and Content Multiplier uploads, voice notes) use OpenAI’s GPT-Transcribe, with Scribe v2 as the backup; the Clipper, meeting minutes and podcast uploads still use OpenAI’s Whisper for now. No new processor is involved.
Cloud film renders now run fully under Fal’s DPA
Movie Director’s cloud film renders no longer use two Fal utilities that Fal hadn’t placed under its data processing addendum. Each scene is now prepared and the finished film checked on our own servers, and the scenes are joined with their soundtrack by Fal’s FFmpeg compose, which Fal’s DPA covers. No new processor is involved.
Background removal by Bria AI, upscaling by SeedVR2, and where each developer is
Background removal now uses Bria AI’s RMBG 2.0 (Bria is in Tel Aviv, Israel), reached through Fal’s partner API, so the photo being cut out goes to Bria. It replaces an open model Fal ran on its own servers. The Image Upscaler now uses SeedVR2, an open model Fal runs itself. The model developers table now also shows where each developer is.
New speech models through Fal, and a training column
Our newer tools narrate with xAI Text to Speech and ElevenLabs’ Eleven Turbo v2.5, and transcribe with ElevenLabs’ Scribe, all run through Fal; they receive the text to be spoken or the audio to be transcribed. The lists now also show whether each AI provider’s terms let it train on content it receives.
Processor list published
The current list of processors and model developers was published with our updated legal documents.