Privacy Policy
Reason to Manim turns a math question you type into a short explainer video. Doing that means your question has to leave your browser and be processed by machines we rent from other companies. This page explains exactly what is collected, who touches it, and for how long — in plain language rather than legal boilerplate.
- The short version
- What we collect
- Why we collect it
- Who else processes it
- International transfers
- How long we keep it
- Your rights
- Children
- Security
- Changes
- Contact
1. The short version
| What | Why | How long |
|---|---|---|
| The question or topic you type | To work out the answer and plan the video, and to keep it in your history afterwards | Until you delete that job or your account |
| The video we render, and the explanation we generated | To show and replay it for you | Video: 30 days. Explanation: with the job record, until you delete it |
| Your IP address and browser details | To stop abuse and keep the service running | Up to 30 days |
| Your account details (email, name, sign-in method), held by our authentication provider Clerk | To let you sign in and tie your videos to you | Until you delete your account |
| A job reference, in your own browser | To let you reopen your most recent video | Until you clear your browser storage |
| Anonymous metrics about each run — how long it took, how many tokens it used, which model, and what it cost (no question text, no account or job reference) | To see what is slow, what fails, and what each video costs us to make | Kept indefinitely |
You need a free account to generate videos. Accounts are provided by Clerk, which receives your email address and name when you sign up. We do not ask for payment details, we do not use advertising or cross-site tracking cookies, and we do not sell personal information.
2. What we collect
2.1 What you submit
The text you type into the prompt box, and nothing else. The box is a single free-text field; there is no file upload, no microphone access, and no camera access. Please do not type personal information about yourself or anyone else into it — it becomes part of your job record and is processed by the systems described in section 4.
We store your question together with your job so that you can find it again in your history, and so that we can work out what went wrong when a run fails. The job record — your question, the explanation we generated, the run status, and the metrics described in section 2.4 — is kept until you delete that job or your account. The rendered video is not kept that long. It is deleted from object storage 30 days after it is created, and once the video is gone that entry stops appearing in your history; what you see there is therefore the finished videos from the last 30 days, plus any runs that failed (those never had a video to lose, so they stay). We keep records older than that to diagnose failures and understand how the service behaves. Deleting a single job, or your account, removes the record immediately, and you can ask us to delete a specific record at any time — see section 6.
2.2 Information we receive from your sign-in provider
Accounts are run by Clerk, and you can sign in with an email address or with a third-party account such as Google. When you use a third-party account, that provider passes us the minimum needed to create and secure your account: typically your email address, your name, an identifier for that account, and whether the email has been verified. We never receive your password for a third-party service, and we cannot see anything in that account beyond what the sign-in consent screen tells you is shared. Which sign-in methods exist is configured in Clerk's dashboard, not in our code.
If you would rather not share a third-party profile, sign in with an email address instead. Either way, what we store on our side is the same: the Clerk user identifier that your videos are attached to.
2.3 What the service generates
The explanation text and the finished video file. These are stored so that the video keeps playing when you refresh the page, close the tab, or open your link on another device. We also retain diagnostic records of each model call: response text, tool arguments (which can include animation source code), tool results, timing and usage. These records are associated with your job and are deleted when you delete the job or your account. Large diagnostic payloads may be truncated.
2.4 Technical information
Like every website, our infrastructure provider records the request metadata needed to serve traffic and block abuse: IP address, browser user-agent string, the page or endpoint requested, and a timestamp. We also keep a short-lived per-IP counter so that one visitor cannot consume the whole service: the current allowance is a small number of generations per network every ten minutes, and the counter is discarded when the window closes.
Making a video also costs money and takes minutes, so we record what each run took: the model and reasoning setting used, how long the run spent in each stage, how long the model and the renderer each took, how many tokens were consumed, and what the run cost. This is part of your job record (see section 6) and is deleted with it. Separately we keep a copy of those numbers in our own database with the question text, your account identifier, and the job reference removed — so the copy cannot be traced back to you — and that anonymous copy is the one we use to see which stages are slow, how often runs fail, and what the service costs to operate. Because nothing in it identifies you, it is not covered by the deletion promise in section 6 and we keep it for as long as it is useful for that analysis.
2.5 Storage in your browser
Our own code sets no cookies. We keep two localStorage entries — the reference to your most recent video, and a record that you have seen the cookie notice. There is no language setting: the interface is English only. Signing in additionally requires Clerk, which sets the strictly necessary authentication cookies. Every entry, with its class, purpose and lifetime, is listed in the Cookie Policy.
2.6 What we do not collect
- No payment or financial data. (If we introduce paid plans we will update this policy first — see section 10.)
- No advertising identifiers, no social media pixels, no cross-site tracking.
- No third-party analytics, tracking, or advertising product of any kind, and nothing that follows you from site to site. The only measurement we keep is our own anonymous service metrics described in section 2.4 — they contain no question text and no account or job reference, and we do not use them to build a profile of you. If we ever add a product that tracks visitors, we will ask for your consent first and update this page and the Cookie Policy before it goes live.
3. Why we process it, and our legal basis
Where the EU or UK GDPR applies, we rely on the following legal bases:
| Purpose | Legal basis |
|---|---|
| Generating and delivering the video you asked for | Performance of a contract (Art. 6(1)(b)) |
| Creating your account, checking that you are signed in, and keeping it secure | Performance of a contract (Art. 6(1)(b)) |
| Keeping your question in your history so you can find it again | Performance of a contract (Art. 6(1)(b)) — it is part of the service you signed up for |
| Keeping a record of a run after its video has expired, so we can diagnose failures and understand how the service behaves | Legitimate interests (Art. 6(1)(f)) — our interest in being able to fix what breaks. You can delete any single job or your whole account at any time, which is what keeps this proportionate |
| Rate limiting, abuse prevention, and keeping the service available | Legitimate interests (Art. 6(1)(f)) — our interest in not having the service drained or attacked |
| Complying with legal obligations, or defending legal claims | Legal obligation (Art. 6(1)(c)) / legitimate interests (Art. 6(1)(f)) |
We do not use your submissions to train machine-learning models, and we do not build advertising profiles from them.
4. Who else processes it
We are a small operation and rely on the following processors. Each one receives only what it needs:
| Provider | Role | What it receives |
|---|---|---|
| Clerk, Inc. | Authentication: accounts, sign-in, and session management | Your email address, name, and the sign-in method you choose; session tokens |
| Cloudflare, Inc. | Hosting, CDN, durable job state, video object storage, database holding the anonymous service metrics | The full request, your IP address, the generated video and code, and the anonymous run metrics described in section 2.4 |
| DeepSeek (Hangzhou DeepSeek Artificial Intelligence Co., Ltd.) | Language model that solves the problem, plans the video, and writes the animation code | Your question text, the conversation so far, and the generated code |
| Our rendering service (self-hosted ManimGX) | Runs the generated animation code to produce the video file | The generated animation code |
Rendering happens on servers we operate ourselves. If we ever move it to a third-party provider, we will update this policy and the sub-processor list before doing so.
We may also disclose information if we are legally required to, or to protect the rights and safety of users, the public, or ourselves. If the service is ever sold or merged, your information may transfer to the buyer under the same commitments in this policy.
A current list of sub-processors, and the data processing agreement we hold with each of them, is available on request from privacy@reason2m.com.
5. International transfers
We operate from the People's Republic of China and our processors store and process data in several countries, including the People's Republic of China, the United States, and the European Union. Where personal data leaves the EEA, the UK, or Switzerland, we rely on the European Commission's Standard Contractual Clauses (and the UK Addendum where relevant) together with the processors named above, and we assess each transfer for the risk of government access. You can request a copy of the relevant safeguards from privacy@reason2m.com.
6. How long we keep it
- Job records — your question, the explanation we generated, the reference to your video, the status of the run, and the run metrics described in section 2.4 — are kept until you delete that job or your account. There is no automatic expiry.
- Rendered videos are deleted automatically from object storage 30 days after they are created. A job record outlives its video, and once the video is gone the entry no longer appears in your history — but the record itself stays until you delete it or your account. Runs that failed never had a video, so they stay listed: that way you can see what you asked and that it did not work. Older records that are no longer listed are still stored, which is why you can email us to have a specific one deleted.
- Account details (email, name, sign-in method) are held by Clerk for as long as your account exists. Deleting your account removes them from Clerk and also deletes your job records and rendered videos — that happens at the moment of deletion, not on any timer.
- Rate-limit counters are discarded as soon as the counting window closes (currently ten minutes).
- Infrastructure logs held by Cloudflare are retained for up to 30 days according to their configuration.
- Anonymous service metrics (section 2.4) are kept indefinitely. They hold no question text, no account identifier, and no job reference, so they are not personal data and there is nothing in them to expire or to delete — which is also why deleting your account does not affect them.
A video link is public: anyone who has it can open the video while it exists, and it stops working once the video expires after about 30 days. Do not share a link to a video you want to keep private. Re-run the question to generate a fresh one.
If you want your data removed sooner than any of the above, delete the job from your history or delete your account — both take effect immediately. You can also email privacy@reason2m.com with the link or the job reference from your browser's address bar and we will delete it. We may keep a minimal record where the law requires us to.
7. Your rights
If you are in the EEA, the UK, or Switzerland, you have the right to: access your personal data; correct it; erase it; restrict or object to our processing; receive it in a portable format; withdraw consent where we relied on it; and lodge a complaint with your local supervisory authority.
If you are a California resident, you have the right to know what we collect, to request deletion, to correct inaccurate information, and not to be discriminated against for exercising those rights. We do not sell or share personal information as those terms are defined by the CCPA/CPRA, and we have not done so in the preceding twelve months.
Residents of other US states with comprehensive privacy laws (including Colorado, Connecticut, Virginia, Utah, and Texas) have equivalent rights, including appeal of our decision — reply to our decision email and we will escalate it to a human reviewer.
If you are located in the People's Republic of China, you additionally have the rights under the Personal Information Protection Law (PIPL), including the rights to know, access, correct, and delete your personal information, to request its transfer, and to withdraw consent.
To exercise any of these rights, email privacy@reason2m.com. We will not charge you for a reasonable request, and we aim to respond within 30 days. Where we relied on your consent, you may withdraw it at any time; withdrawing does not affect the lawfulness of what we did before you withdrew.
Deleting your account is available inside the product and not only by email: the account menu in the app removes your Clerk account, and we then delete your job records and the videos we rendered for you from our storage. You do not have to email us for this to happen. If you delete the account somewhere else — for example from Clerk's own account menu — the same cleanup runs when Clerk notifies us. Either way you can ask us to confirm what was removed.
We do not carry out automated decision-making or profiling that produces legal or similarly significant effects about you.
8. Children
Reason to Manim is not directed at children under 13, and we do not knowingly collect personal information from them. Because parts of the world require a higher age of digital consent, we set the minimum age at 16 for users in the EEA and the UK, and at 13 elsewhere (see the Terms of Service). We do not ask for an age and we do not use your question for anything other than producing your video — but if you believe a child has used the service, contact privacy@reason2m.com and we will delete the associated data.
9. Security
The service runs over TLS. Generated animation code is executed only inside an isolated third-party sandbox, never on our own infrastructure and never on your device. Access to production secrets is limited to the operators of the service. No system is perfectly secure, so please treat anything you type as potentially retained, and never submit confidential or sensitive information.
10. Changes
If we change how we handle personal data we will update this page and move the date at the top. Material changes — such as introducing analytics or payments — will be announced on the service itself before they take effect.
11. Contact
The data controller is Reason to Manim, operated by an individual developer, reachable at privacy@reason2m.com.
Privacy questions, data requests, and sub-processor enquiries: privacy@reason2m.com.
General support: support@reason2m.com.