Privacy policy

Version 5, 5 November 2026. Terranext SIA, Riga.

What changed in version 5: studios and companies we write to first (where we find them, what we keep, how to stop us), the receipt for a support request, what our fixing agents read, and the AI that drafts our answers to support requests from 7 December 2026.

Clarified on 10 November 2026: the YouTube data we read (your playlists' names and the caption languages) is now listed. What we do with it has not changed.

What changed in version 4: the data we process for studios, the safeguard for each company outside the EU, and invited members.

What changed in version 3: sign-in codes and links, and the site's forms (the early-access form is now the contact form).

What changed in version 2: the free first episode, the site's early-access form, support requests, stock images, TikTok, voiceprints on your computer, what reaches other companies directly from the app, fixed retention periods, how to download your data or close the account, and Google's required wording for YouTube.

Studiotake is a desktop application and a hosted service for podcast studios, made by Terranext SIA, Pulkveža Brieža iela 2, Riga, LV-1010, Latvia, VAT LV40203700622. This policy says what we hold about you, why, and for how long. It applies to the app, to api.studiotake.app, to studiotake.app, and to the messages we send to studios and companies we write to first. Terranext SIA is the controller of the data listed under "What the service holds" and under "Studios and companies we write to first". For the text sent to the AI provider, the videos waiting to be posted, and other people's data in a support request, we act for the studio, as its processor, under the Data Processing Terms (studiotake.app/dpa).

What stays on your computer

Your recordings, transcripts, cut lists, edits, renders, exports, music, brand kits and the shows' settings live on your computer or your own drives. The sound analysis (who speaks when, noise, levels) and the transcription run on your computer. Your recordings are never uploaded to Studiotake.

Three things leave your computer through Studiotake's server, and only when you use them: the text the AI works on passes through it to the AI provider and is not stored; a finished video or image you schedule for Instagram, Facebook or TikTok is stored on it until it is posted; and support requests you send. The sections below say exactly what.

The files on your computer are yours, and so are the decisions about them: for what you record and keep there, you (the studio) are responsible under the data protection law that applies to you, not us.

Voices the app remembers (voiceprints)

To suggest who is speaking in a new episode, the app can remember the voices of a show's hosts and co-hosts. A remembered voice is a set of numbers computed from the recording (a speaker embedding). It can identify a person's voice, so treat it as biometric-type data.

• It is kept only on your computer, in the show's folder (the file voices.json). It is never sent to Studiotake, to the AI provider or anywhere else. Studiotake does not receive it and cannot see it.
• It is learned only for a host or co-host, and only from a name you picked yourself for that voice. A guest's voice is never kept.
• You can forget one voice or all of them at any time in the show's settings ("Studiotake remembers your hosts' voices on this computer · Forget"). Deleting the show's folder deletes them too.

Because this happens on your computer and under your control, the studio is responsible for telling the people it records, and for getting their consent where the law requires it, before it lets the app remember their voices. If you do not want a voice kept, do not name it, or forget it afterwards.

What the service holds

Of the platforms' tokens, the long-lived ones (and TikTok's daily ones) are encrypted with AES-256-GCM; a YouTube access token, which lasts an hour, is kept as it is until it ends. We hold no transcripts and no card numbers. Deleting the account deletes all of the above (see "Your rights").

What the app sends to the AI provider

To write the cut, the titles, the posts and the other texts, the app sends the AI provider (Anthropic, Claude) the transcript text a step needs, with the speakers' names, the show's instructions for that step, and, when you use the studio assistant, what you tell it with the plan and transcript lines it reads. It never sends audio, video, files, voiceprints or keys.

It goes through Studiotake's server to Anthropic. Our server does not store it: it keeps only the counts and cost listed above. Anthropic keeps API requests for up to 30 days under its commercial terms (longer only for a request it flags for a policy violation), and does not train its models on them.

What the app fetches or sends directly

Some things go from the app on your computer straight to another company, not through Studiotake:

• Models. At set-up, the app downloads its transcription model from Hugging Face and its voice models from GitHub. They see your computer's internet address, as for any download.
• Your brand. If you give the app your website, it reads that site (the home page and at most one more page, its styles and images) from your computer, and downloads the open fonts it names from Google Fonts.
• YouTube uploads. The video goes from your computer to YouTube (see "Google and YouTube").
• Integrations you add yourself. If you connect Epidemic Sound or another MCP server in Settings, the app and the studio assistant send that service the searches and requests you make with it, with your own token. That is between you and that service.

The site's forms

The forms on studiotake.app (the contact form on the home page and on /pro, and the application form on /partners) collect your name, your e-mail address, what you make, the form you used and your message. We store it, send a copy to our mailbox (contact@studiotake.app) and send you a short confirmation. Our team keeps a note of where things stand with you (replied, in talks, a follow-up date) and whether you have since made a Studiotake account. We use it only to answer you. We do not add you to a newsletter. We delete it 12 months after the last contact, sooner if you ask; it is also deleted when a Studiotake account with that address is deleted.

The site's pages load their typefaces from Google Fonts, so Google sees your internet address when you open them. The site sets no cookies of its own; connecting a channel sets one short-lived cookie on api.studiotake.app, used only to finish that connection.

How we answer support requests

When you send a support request from the app, our server e-mails you an automatic receipt with its reference, so you know it arrived. It does not repeat anything you wrote.

An AI already helps us fix faults. Our fixing agents run with Anthropic's Claude. They read a support request as it reached our server: your words, what you left ticked (with keys, home folders and the episode's words already taken out by the app), the plan or screenshot if you ticked them, and your studio's name. They use it to reproduce the fault on test material, never on your episodes, and to propose a fix, which a person at Studiotake reviews before it ships. They may post a short status in the app, such as "Looking at it", or ask you a question there when they need more detail. They never e-mail you.

From 7 December 2026, the AI also drafts our answers: it says what kind of problem it is and writes a draft, and a person at Studiotake reads every draft, changes it where needed and sends it. If we ever let the AI answer some kinds of question by itself, the reply will say so, and we will change this policy first.

Studios and companies we write to first

We write to podcast studios, production companies, producers and independent podcasters, to tell them about Studiotake, even when they have not written to us. If that is you, this is what we hold and why. Where this section says "your organisation", for an independent podcaster it means you and your show.

• What we keep. Your name and role; your organisation, its city and country; a work e-mail address or contact page that you or your organisation publish; facts about your organisation's work (how it records, the shows it makes, whether it offers editing), each with the public page it comes from; our messages to you and your answers; and where things stand (interested, not now, stop).
• Where it comes from. Your organisation's website and social pages, public professional profiles such as LinkedIn, published episodes and podcast directories, business directories and job adverts. We use only what people and organisations have made public, and we do not guess e-mail addresses.
• Who does the work. Our team, helped by AI assistants that read public pages and draft messages: Anthropic's Claude and SpaceXAI's Grok Bot. A person at Studiotake decides whom we write to and approves every message. We score how well Studiotake fits an organisation only to decide whom to write to; the score has no other effect.
• Why we may. Our legitimate interest in offering a product to businesses it may help. We write only to work addresses, about your work, a few times at most, and we stop at your first word.
• What our messages say. Who we are, where we found your details, and how to stop: reply "stop" or "no", or write to contact@studiotake.app. Our messages may come from an address at studiotake.app or at another domain Terranext SIA owns for them; each names Studiotake and links to this page.
• How long. If we find your details but have not written to you within one month, we delete them. If you do not answer, we delete your details 12 months after our last message. If you say "not now", we keep them until the time you suggest and delete them 12 months after it at the latest. If you ask us to stop, we stop writing at once and delete your details within 30 days, from our sheet, our mailboxes and lemlist. We keep only your e-mail address (or your organisation's web address, if you ask for the whole organisation) on a do-not-contact list, so that we never write again. If you become a customer, the rest of this policy applies.
• The services involved. Our mailboxes and the sheet where we keep these details are on Google Workspace. We may send these messages and receive the answers through lemlist, under its data processing agreement. LinkedIn sees what we send there.

You can object at any time, without giving a reason, and we stop. You can also ask what we hold about you, have it corrected or deleted, or ask us to stop using it while a question is settled, at contact@studiotake.app. We answer within one month. You can complain to your data protection authority (in Latvia, the Data State Inspectorate).

Who else sees data

• Anthropic (Claude, United States): receives the text described under "What the app sends to the AI provider", support requests as described under "How we answer support requests", and what we gather about studios and companies we write to first.
• SpaceXAI (Grok Bot, United States): the AI assistant we are testing for research and drafts handles what we gather about studios and companies we write to first.
• lemlist (France): when we use it, sends our messages to studios and companies we write to first, and receives their answers.
• LinkedIn (LinkedIn Ireland Unlimited Company, Ireland): sees the connection requests and messages we send there.
• Hetzner (Germany and Finland) hosts the service and stores its encrypted backups.
• Resend sends our e-mails (sign-in codes and links, invitations, receipts of your messages).
• Google Workspace holds our mailboxes: contact@studiotake.app, where copies of your messages and notices about new subscriptions arrive, and the ones we write to studios and companies from.
• Stripe handles payment. Stripe collects your card, billing address and VAT number, and keeps the invoices as accounting law requires; we hold no card numbers.
• Shutterstock receives the search words for stock images, and licenses the images under Studiotake's account. It receives no personal data from us.
• healthchecks.io receives a heartbeat from our server, with no personal data.
• Google Fonts, Hugging Face and GitHub, as described above.
• Google (YouTube), Meta (Instagram, Facebook) and TikTok, when you connect them, as described in the sections below.
• Nobody else. We do not sell data or share it for advertising.

Copies of e-mails already sent stay with Resend under its own retention, and in our mailbox until we delete them; closing the account does not reach them, but we delete the mailbox's copies when you ask.

Our contract with Anthropic is with Anthropic Ireland, Limited; the AI requests are processed in the United States under the EU Standard Contractual Clauses in Anthropic's data processing terms. Resend, Google and Stripe are certified under the EU-US Data Privacy Framework, with the Standard Contractual Clauses as a fallback. Hetzner keeps everything in the EU. SpaceXAI's Grok Bot, which we are testing, runs in the United States under SpaceXAI's own terms. You can ask us for a copy of these safeguards at contact@studiotake.app. The companies that process data for studios are also listed at studiotake.app/subprocessors.

Google and YouTube

When you connect a YouTube channel, you grant Studiotake the permissions Google lists on its consent page: to upload videos (youtube.upload) and to manage your videos' captions, thumbnails and playlists (youtube.force-ssl). Our server keeps the channel's id and name, the long-lived token Google issues (encrypted) and, for an hour at a time, the access token made from it. For an upload, the app on your computer gets that access token from our server and sends the video straight to YouTube: the video does not pass through our server.

We use this access only to publish what you approved in the app, with its title, description, chapters, captions and thumbnail, to add it to your playlists, and to read your channel's name, the names of your playlists, the caption languages and the state of what we published. We do not use YouTube data for advertising, we do not sell it, and we do not let anyone read it. We do not use data from Google APIs to develop, improve or train generalized AI or machine-learning models. YouTube data is not sent to the AI provider.

Studiotake stores and accesses information on your device: the app keeps the ids, titles and publishing state of the videos it uploaded in your show's folder on your computer, and reads them back to show and update your schedule.

Studiotake's use and transfer of information received from Google APIs adheres to the Google API Services User Data Policy, including the Limited Use requirements.

Studiotake uses YouTube API Services. By connecting a channel you agree to the YouTube Terms of Service, and Google's Privacy Policy applies to what Google does with your data.

To stop it: press Disconnect on the Accounts screen of the app, which revokes the token at Google and deletes the channel's id, name and tokens from our server at once. You can also remove Studiotake at any time in your Google Account's security settings (myaccount.google.com/permissions); our server checks each connection with Google once a day, and deletes the same data as soon as Google says the permission is gone.

Meta: Instagram and Facebook

When a studio connects an Instagram professional account (Instagram Login) or a Facebook Page (Facebook Login for Business), Studiotake keeps:

• the Instagram account's id and username, or the Page's id and name;
• the id Meta gives the person who connected it for Studiotake's app;
• the access token Meta issues (for Instagram, a token that lasts 60 days and that we renew while the connection is in use; for a Page, the Page's token), encrypted with AES-256-GCM on our server. It never leaves the server: the app on your computer does not receive it.

Studiotake asks Meta for these permissions: for Instagram, instagram_business_basic and instagram_business_content_publish; for Facebook, pages_show_list, pages_read_engagement and pages_manage_posts, and optionally business_management (to list Pages that belong to a business). We use them only to publish what the studio scheduled in the app, at the time it chose (a Reel, a video or an image with its caption), to list and name the accounts and Pages you can choose, and to read the state of what we published (its link, and Instagram's daily publishing limit). We do not read your messages, followers, comments or insights, and we do not use Meta data for advertising or share it.

To stop it: press Disconnect on the Accounts screen of the app, which deletes the account's id, name and token from our server and the files waiting to be posted through it, and, for Facebook, removes Studiotake's permission at Meta. You can also remove Studiotake in Instagram (Settings → Apps and websites) or Facebook (Settings → Business integrations); we then delete the token as soon as Meta tells us. To have everything deleted, follow the data deletion instructions.

TikTok

When a studio connects a TikTok account (Login Kit), Studiotake keeps the account's open id, its display name and username, and the tokens TikTok issues, encrypted on our server. We use the permissions user.info.basic and video.publish only to read the account's name and its posting options, to post what the studio scheduled with the choices made for that post, and to read whether it was posted. The file is deleted from our server once it is posted, fails or is cancelled. Posting to TikTok also follows TikTok's own terms.

To stop it: press Disconnect on the Accounts screen of the app, which revokes the tokens at TikTok and deletes the account's open id, name and tokens from our server, with any file still waiting to be posted through it. You can also remove Studiotake in TikTok (Settings and privacy → Security → Manage app permissions); TikTok tells our server, which deletes the same. See the data deletion instructions.

Where

The service runs in the European Union (Helsinki). Backups are encrypted before they leave the server, stored in the EU and kept for up to 12 weeks.

Why we may use it

We use your account, plan, connections, usage and support data to provide the service you signed up for (performing our contract with you); the activity log, sign-in records and access logs to keep accounts and the service secure (our legitimate interest); billing records because tax law requires it; the forms' messages to answer what you asked us; and the details of studios and companies we write to first because of our legitimate interest in telling businesses about Studiotake. Where we write to you first, we have weighed that interest against yours: we use only work details your organisation makes public, write about your work only, say where we found you, and stop at your first word.

Your rights

You can see, correct, export or delete what we hold about you. In the app, Settings → Privacy & data has "Download my data" (everything our server holds about you and your studio, including your messages from the site's forms, as one file) and a way to close the account. You can also write to contact@studiotake.app.

If a studio invited you, it gave us your name and e-mail address. We use them only for your sign-in and the studio's account. The invitation e-mail links to this policy.

Closing the account takes effect after 30 days, so you can change your mind; then the data above is deleted. A studio with a running subscription is deleted once the subscription ends. Invoices stay with Stripe as accounting law requires; where a studio's records still name a person, that person is anonymised instead. Backups age out within 12 weeks.

You can disconnect any channel at any time in the app, which revokes the permission at the platform too, and you can also revoke it from the platform's own account settings. Under the GDPR you may also object to processing based on our legitimate interest, and complain to your data protection authority (in Latvia, the Data State Inspectorate). If we wrote to you first, an objection always stops us (see "Studios and companies we write to first").

Changes

We will announce changes to this policy by e-mail to account owners before they apply.

Contact: contact@studiotake.app · Terranext SIA, Pulkveža Brieža iela 2, Riga, LV-1010, Latvia