Privacy Policy
Last updated:
This describes what Mealbot actually does with your data — written from the code, not from a template. Where something is imperfect, it says so.
The short version. We use three cookies, all of them needed to keep you signed in. We run no analytics, no advertising and no third-party trackers of any kind. We never see your card details. But your dietary restrictions and allergies are sent to Google, and sometimes to OpenAI, to generate your meal plans — that is how the product works, and it is the most important thing on this page. We use them only after you have explicitly consented, and you can withdraw that consent in Settings.
Who is responsible
Mealbot is run by František Bláha, Jana Černého 489/2b, Věkoše, 503 41 Hradec Králové, Czech Republic. IČO 22059946. Not registered for VAT (neplátce DPH). We are the data controller for the personal data described here.
For anything on this page, including any request below, email info@trymealbot.com.
What we collect
Your account
Your email address and a hashed password (bcrypt — we never store or log the password itself). Plus the state we need to run your account: when you registered, whether your email is confirmed, and your subscription status.
What you create in the app
Your cooking preferences (country, language, units, whether to include spices), your fridge contents and pantry staples, your meal plans and recipes, and when you marked a meal as cooked. Free-text fields you fill in — taste preferences, ingredients to avoid, notes — are stored as you wrote them. Please put health details only in the diet, allergy and ingredients-to-avoid fields, which your consent covers (see below), not in taste preferences, notes or fridge items.
Dietary restrictions and allergies
This is health-related information, and in some cases (halal, kosher) it can reveal religious belief. We treat it as the most sensitive thing we hold. It is covered separately below because of where it goes.
We use your diets, allergies and ingredients to avoid only with your explicit consent: a separate checkbox, unticked, that appears the first time you add any of them, before anything is sent. Giving them is optional. Without them, and without your consent, Mealbot still generates plans, but cannot take them into account. You can withdraw your consent at any time in Settings; we then remove them from what we stored with your saved plans, while the meals already generated stay as they are.
If you plan food for other people, such as your household or a child, only enter their diets and allergies if they agree, or if you are their parent or guardian. We store them as part of your plans, without names. They can exercise their rights through you, or by emailing us. Someone who shares your household with an account of their own saves their diets and allergies under their own consent (see "Sharing a household").
Payment
We never receive your card details. Checkout is hosted by Stripe and card data never touches our servers. Stripe collects your name, billing address, card and any VAT ID directly. We store only identifiers and status — your Stripe customer and subscription IDs, whether your subscription is active, and when the period ends.
For each payment we keep an invoice record: amount, currency, your country, and whether it was a business purchase. We need this for VAT.
We also store a one-way, keyed fingerprint derived from your card by Stripe — not the card number, and not reversible. It exists only to stop the same card opening unlimited free trials.
Support and access requests
If you send feedback, we store your message. If you request access from the landing page while sign-ups are closed, we store the email address and message you submit — including any dietary details you choose to mention there. Please do not put health details in either; if you do, we use them only to deal with your message.
Technical
Login sessions (a hash of the session token, timestamps, and your browser's user-agent string). Counts of AI tokens used, to enforce fair-use limits — this records how much was used, never what was in it.
We also keep a copy of what was sent to the AI and what it returned, plus any edits you make to its output, so we can tell where the AI gets things wrong. This copy leaves out your diets, allergies and ingredients to avoid: we remove them before saving it. There is no switch for it in the app, but you can object to it (see below).
Our web server writes standard access logs containing your IP address, the page requested, and your user-agent. We do not put IP addresses in our database and we do not use them for analytics.
These logs are capped in size and rotate automatically — once the cap is reached, the oldest entries are deleted. They are not kept forever. We cap by size rather than by age, so we cannot honestly quote a fixed number of days: at low traffic a line may survive longer, at high traffic much less. The logs never contain the contents of your meal plans, and they never contain sign-in links or password-reset links.
Your dietary data goes to Google, and sometimes to OpenAI
To build a meal plan we send your request to Google's Gemini API, and sometimes to OpenAI's API. That request includes your declared allergens, your diet types (including halal, kosher, diabetic, low-FODMAP and baby-food), your free-text ingredients-to-avoid, your taste preferences, your household size and your fridge contents.
We do not send your email address, your name, or any account identifier. Google and OpenAI receive the content, not who you are.
A request goes to OpenAI when Google's service fails or answers too slowly, in which case Google may have received it too. For a few minutes after Google answers too slowly, new requests may go to OpenAI first. We may also send a share of requests to OpenAI first, to compare which of the two writes better plans.
The same applies to a receipt photo if you use the scanner (the image is sent), and to feedback you submit (the text is sent, to categorise it automatically).
After generation, we run our own deterministic check against your declared allergens and refuse to show a plan that fails it. That is a safety net on top of the model, not a replacement for reading labels.
Google and OpenAI act as our processors for this data, under their data processing terms. The terms of the APIs we use (Google's paid Gemini API and OpenAI's API) forbid both from using it to improve their products or train their models, and we have not opted in to sharing data with OpenAI. The one exception is the copy each keeps for a limited time to detect abuse, which Google may also use to train its abuse filters. OpenAI keeps that copy for up to 30 days, longer only where the law requires it or to protect against harm; Google keeps it for 55 days. If you would rather not have this data processed this way, do not give the consent: Mealbot then generates plans without your diets, allergies and ingredients to avoid, and so cannot take them into account.
Who else receives your data
| Who | What they get | Why |
|---|---|---|
| Google (Gemini API) | Your meal-plan request, including dietary restrictions, fridge contents, receipt images and feedback text. No identity. | Generating plans and recipes |
| OpenAI (API) | The same as Google, for the requests we send to OpenAI (see above). No identity. | Generating plans and recipes |
| Stripe | Your email and internal user ID; they separately collect your name, address, VAT ID and card | Payments and subscriptions |
| Resend | Your email address and the contents of the email we send you | Sending sign-in, password-reset and verification email |
| GitHub | The text of feedback you submit, filed to a private repository under a numeric ID | Tracking reported bugs |
| Hetzner | Everything, as our hosting provider | Running the servers |
| Cloudflare | DNS lookups for our domain | Domain name resolution |
That is the complete list. We do not sell your data, we do not share it for advertising, and no other company receives it.
Servers are hosted with Hetzner Online GmbH in Nuremberg, Germany, so your data is stored in the EU. Google, OpenAI, Stripe, Resend and GitHub may process data outside the EU. Where they do, the transfer relies on the European Commission's adequacy decision for the EU-US Data Privacy Framework, for a company certified under it, or on the EU Standard Contractual Clauses in its data processing terms. OpenAI may process it in the United States, under the EU Standard Contractual Clauses or a European Commission adequacy decision, as its data processing addendum provides.
Sharing a household
A household holds up to six accounts, yours included, and you can invite other people to share yours. Everyone in a household shares one kitchen: the fridge, the pantry staples, the meal plans with their recipes, and the cookbook. Every member can see and change all of it, and members see each other's email addresses. Your cooking preferences and the food you mark as wasted stay yours.
Invites. To invite someone, you enter their email address. An invite works for 48 hours. We keep the address, and a one-way hash of the invite code (never the code itself), until the invite is used or cancelled; an unused one is deleted by the daily clean-up after it expires. Every member of your household can see open invites in Settings, and any of them can cancel one. Only someone signed in with the invited address can use the code.
Joining. When you join a household, your plans that were never confirmed are deleted, and your other plans and recipes, your cookbook, fridge and pantry staples become part of its shared kitchen: every member can see and change them, and they stay with the household if you leave. Your saved diets and allergies and your preferences stay yours.
Your diets and allergies. With the consent described above, the diets, allergies and ingredients to avoid that you save in Settings are shown to the other members, and the household's plans and recipes are generated with them whenever you are among the people eating. Whoever generates a plan chooses who is eating, and can add diets and allergies for that plan alone. When a member edits a meal of a shared plan, Mealbot checks the edited meal against the allergens the plan was made to avoid, including ones added for that plan alone, and warns them when it finds one. Without your current consent, the other members see only that you have saved some, not what they are, and they are not used for new plans. If you withdraw your consent, your diets and allergies saved in Settings are deleted, and every shared plan made with your diets and allergies loses all of its diet and allergy details, the other members' included, while its meals stay.
Leaving. When you leave, you move to a household of your own and take your saved diets and allergies and your preferences with you. The shared kitchen stays with the others. The diets and allergies you saved in Settings are removed from the plans you shared, which keep working for the others; a plan where you added diets or allergies for that plan alone loses all of its diet and allergy details, and says so. That includes plans you brought along when you joined. Your plans that were never confirmed are deleted, and so are the invites you sent. If we deactivate your account, we move you out of the household the same way.
Deleting your account while you share a household deletes your account, your preferences and your plans that were never confirmed. The shared kitchen stays with the other members, and the plans made with your diets or allergies keep their meals but lose all of their diet and allergy details, the other members' included.
Why we are allowed to use your data
The GDPR requires a legal basis for each use. These are ours:
| What we do | Legal basis |
|---|---|
| Running your account, signing you in, your subscription, generating plans and recipes, and emails about your account | Our contract with you (GDPR Art. 6(1)(b)) |
| Using your diets, allergies and ingredients to avoid, including sending them to Google and OpenAI, and showing them to the members of your household | Your explicit consent (Art. 9(2)(a) and Art. 6(1)(a)), which you can withdraw in Settings at any time |
| Sharing your household's fridge, pantry staples, plans and cookbook with its members, and showing them each other's email addresses | Our contract with you (Art. 6(1)(b)), since sharing is what you asked for |
| Keeping the email address of someone you invite, and showing it to your household's members, until the invite is used or cancelled, or until the daily clean-up after it expires | Your and our legitimate interest in letting you invite the people you choose (Art. 6(1)(f)) |
| Invoice records, and your customer record at Stripe | Our obligations under tax and accounting law (Art. 6(1)(c)) |
| The record of your consent and of its withdrawal | Our obligation to be able to show that you consented (Art. 6(1)(c) with Art. 7(1)) |
| The admin log of actions we take on accounts, such as deleting one on your behalf | Our legitimate interest in being able to show what we did and to defend legal claims (Art. 6(1)(f)) |
| Access requests from the landing page | Steps you ask us to take before a contract (Art. 6(1)(b)) |
| The copy of AI requests and your edits to their output | Our legitimate interest in finding and fixing the AI's mistakes (Art. 6(1)(f)) |
| The card fingerprint | Our legitimate interest in stopping one card from opening unlimited free trials (Art. 6(1)(f)) |
| Login sessions, access logs, database backups and AI token counts | Our legitimate interest in keeping Mealbot secure and available and preventing abuse (Art. 6(1)(f)) |
| Feedback and its automatic categorisation | Our legitimate interest in handling your reports and improving Mealbot (Art. 6(1)(f)) |
You need to give us an email address and a password to have an account. Diets, allergies and ingredients to avoid are optional.
We make no decisions about you based solely on automated processing that have legal or similarly significant effects (Art. 22). The fair-use limit and the free-trial check are applied automatically; if you think one was applied to you wrongly, email us and a person will look at it.
Cookies
We use three cookies and they are all strictly necessary, so there is no consent banner to click:
| Cookie | Purpose | Lifetime |
|---|---|---|
mealbot_at | Keeps you signed in | 15 minutes |
mealbot_rt | Renews your session | 30 days |
mealbot_csrf | Blocks cross-site request forgery | 30 days |
If you never sign in, we set no cookies at all. There is no analytics cookie, no advertising cookie and no session-replay script anywhere on this site.
Your browser also stores some things locally so the app works — who you are signed in as, your saved preferences, and your place in a recipe you are cooking. That stays on your device and is cleared when you sign out.
How long we keep things
| What | How long |
|---|---|
| Login sessions | Deleted 7 days after they expire |
| Meal plans you never confirmed | Deleted when you next generate a plan |
| Demo accounts and their data | About 2 hours |
| Database backups | 14 days |
| Your account and everything in it | Until the account is deleted; a shared household's kitchen stays with its other members |
| Household invites | Until used or cancelled; an unused one is deleted by the daily clean-up after it expires (48 hours) |
| The copy of AI requests, and the record of your consent | Until the account is deleted |
| The admin log of actions we take on accounts | No fixed end date, so we can show what we did |
| Invoice records (VAT) | Kept after deletion, as tax law requires |
Your rights — and what we can actually do today
You have the right to access your data, to have it corrected or deleted, to restrict how we use it, to receive it in a portable form, and to object. Where we rely on your consent, you can withdraw it at any time; that does not affect what we did before you withdrew it.
Three of them you can exercise yourself, in Settings, without asking us: change your email address, download your data, and delete your account. For a correction, or to restrict how we use your data, email info@trymealbot.com and we will action it.
Being straight with you about what those two buttons do and do not cover, because a policy that overstates them is worthless:
- The export contains what you put in. Your profile and preferences, your meal plans and their recipes, your cookbook, your fridge and pantry staples, the bug reports you sent us, and your paid invoices. The file names its own omissions, so you can see what is not in it: your password (we only ever store a hash), session tokens, and the model-performance telemetry described above — ask and we will send you that too.
- In a shared household, the export says which parts are shared. It contains the household's plans, fridge, staples and cookbook, and only your own diets and allergies. It leaves out the other members' diets and allergies and what they asked for when making plans; of their unconfirmed plans it includes only those carrying your diets and allergies, with just the dates, the size and your own diets and allergies: the rest is theirs to export. After you leave a household, ask us if you want what you typed into its plans.
- Deletion is immediate and cannot be undone. There is no grace period and no recovery. If you have a subscription it is cancelled at the same moment, and the rest of the period you have paid for is not refunded — so export anything you want to keep first.
- Deletion is not absolute. Your invoice records are kept, with your account no longer attached, because tax law requires it; your customer record at Stripe and any bug report already filed with our issue tracker are not removed automatically; and you will remain in database backups for up to 14 days. Any action we have taken on your account, such as deleting it on your behalf, stays in our internal admin log with your email address as the record that we did; deleting the account yourself adds no such entry.
Your right to object. Where we rely on our legitimate interests (see "Why we are allowed to use your data"), you can object at any time, on grounds relating to your particular situation, by emailing info@trymealbot.com. We then stop, unless we have compelling legitimate grounds that override your interests, or need the data to establish, exercise or defend legal claims.
You can also complain to the Czech data protection authority (Úřad pro ochranu osobních údajů).
Security
Passwords are hashed with bcrypt. Session, reset and verification tokens are stored only as SHA-256 hashes, are single-use, and expire. The site is HTTPS-only. Cookies are HttpOnly, Secure and SameSite. We use a strict Content-Security-Policy that blocks third-party scripts. Card data never reaches our servers.
No system is perfect. If you find a security problem, please email us — we would much rather hear it from you.
Children
Mealbot is not intended for children under 16. The baby-food option is for adults planning food for an infant; we do not knowingly collect data from children.
Changes
If we change this policy we will update the date at the top. If a change materially affects how we use your data, we will email you about it.