Privacy policy
Status: DRAFT — not reviewed by counsel
Last updated: 2026-07-07
BestPrompter (“we”, “us”) is a prompt-optimization service at bestprompter.com. You give us a prompt or a description of what you need; we return an improved prompt, sometimes with evaluation evidence. This policy explains what data we collect, how we use it, and the choices you have. We have tried to keep it plain and specific.
The short version
- We do not train on your content by default. Using your prompts or results to improve our system is opt-in, and opted-in data is anonymized first.
- Your prompts are sent only to the model providers we list in the product — nowhere else. [TBD before public launch: final provider list, DPA status, and provider no-training/retention commitments.]
- Free-tier auto-run results are published anonymously and listed in our public Explore gallery by default. Paid-tier results are private by default.
- Eligible deletions keep the deleted content recoverable for 30 days, then it is purged. Free-tier public results are removed through account deletion only; paid/manual public pages are taken down immediately when revoked through available controls.
What we collect
Account data. Email address, password hash (if you register with email and password), display name, and avatar. If you sign in with Google or GitHub, we receive your basic profile (name, email, avatar) from that provider.
Your content. The prompts and task descriptions you submit, any attachments you provide, the optimized prompts and results we generate for you, your reviews and feedback on results, your bookmarks, and any public pages you create.
Usage data. For every model call we log the provider, model, token counts, and cost. We also record product events such as runs started, results delivered, and shares created. This supports your run limits, billing, and abuse prevention.
Technical data. Standard server logs (IP address, browser type, timestamps) and the session cookie described below.
Analytics. [TBD before public launch: analytics provider and configuration.] The current proposal is Cloudflare Web Analytics because it is cookieless and aggregate-only, but this draft should not be treated as a final analytics commitment until the Marketing implementation and privacy review confirm it.
How we use your data
- To run the service: optimize your prompts, deliver results, maintain your history and bookmarks, and generate public pages you have chosen (or defaulted) to publish.
- To operate accounts, enforce run limits, and calculate usage for billing.
- To keep the service safe: rate limiting, spend caps, abuse detection, and content guardrails once the required server-side controls are implemented for public launch.
- To improve the service — only in an aggregate, methodology-level way by default. Our system learns from signals like which optimization methods and evaluation setups worked, not from your raw prompt text. Your prompts, outputs, and results are not used to train or improve our models, scorers, or routing unless you explicitly opt in. If you opt in, the data is anonymized before use.
We do not sell your data or use it for third-party advertising.
Third-party model providers
To optimize your prompt, we forward it (and related context) to large language model providers using our own API keys. We only use the providers disclosed in the product; the current provider list must be shown in your account settings and on the run page before public launch. Providers process your prompts to generate responses; their handling is governed by their own terms.
- [TBD: whether provider agreements include zero-retention / no-training commitments, and DPA status with each provider.]
Public pages and the Explore gallery
- Free tier: results from free auto-runs are published as anonymous public pages by default and listed in the public Explore gallery. Public pages never show your identity, your private history, your reviews, or internal evidence traces — only the optimized prompt, what it does, and its scores. In the current version, free users cannot directly revoke the page or delete the underlying session as a removal path; deleting your account is the user-controlled way to remove free auto-run public pages.
- Paid tiers: results are private by default. Nothing becomes public unless you explicitly share it. Manually shared pages are unlisted (reachable only by link) unless you choose otherwise.
- Removal and revocation are immediate where supported. When a public page is revoked or removed through an available control — such as paid/manual revoke, regeneration of a manual link, eligible session deletion, account deletion, or platform takedown — the URL returns “410 Gone”, the page is removed from Explore and search entries, and cached copies and generated preview images are purged.
Cookies and sessions
We use a single first-party session cookie to keep you signed in. It is httpOnly and secure, meaning scripts on the page cannot read it and it is only sent over encrypted connections. Signing in with Google or GitHub uses those providers’ OAuth flows; they may set their own cookies during sign-in under their own policies. We do not use third-party advertising cookies.
Retention and deletion
- While your account is active, we store your sessions, runs, results, reviews, bookmarks, publications, and usage events so that history, bookmarks, sharing, and billing work.
- When you delete eligible content (for example, a bookmark, a paid/private session where the product provides deletion controls, or your account), it is soft-deleted with a 30-day recovery window, then purged from our active systems. Free auto-run public pages are not individually removable by session deletion in the current version; account deletion is the user-controlled removal path.
- When you delete your account, your sessions, runs, reviews, and bookmarks are soft-deleted on the same 30-day schedule, any active public pages are revoked immediately, generated public assets are scheduled for deletion, and your usage events are anonymized (the link to your user ID is removed) so aggregate statistics no longer identify you.
- [TBD: retention period for server logs and backups.]
Where your data is stored
- [TBD: hosting region(s) and data residency commitments.]
Your rights
You can access and update your profile in settings, and view your full history in the product. You may request:
- Access / export — a copy of your account data and content.
- Deletion — of eligible individual content or your whole account, subject to the 30-day recovery window described above. Free auto-run public pages are removed through account deletion only in the current version.
- Opt-out — you are opted out of content-based training by default; if you opted in, you can withdraw at any time (this stops future use of your content).
To exercise these rights, contact us at privacy@bestprompter.com or the address below. Depending on where you live, you may have additional statutory rights (e.g., under GDPR or CCPA). [TBD: lawful-basis mapping for EU users; whether a designated DPO is appointed.]
Children
BestPrompter is not directed at children under 16, and we do not knowingly collect data from them. If you believe a child under 16 has an account, contact us and we will delete it.
Changes to this policy
We may update this policy as the product evolves. Material changes will be announced in the product and the “Last updated” date will change. Continued use after a change takes effect means you accept the updated policy.
Contact
- Privacy contact: privacy@bestprompter.com
- BestPrompter, 1458 3rd Street, San Francisco, California 94158, USA · +1 347 688 5021
(Operating name; replace with the registered legal entity if the service is incorporated separately.)